fix: fix tests

* feat(entity-count): add base structure to show count

* feat(entity-count): add count for clusters

* feat(entity-count): add count for namespaces

* chore(infra-monitoring): move component to be under components folder

* fix(pr): address comments

.github/workflows/integrationci.yaml

@@ -53,12 +53,12 @@ jobs:
           - queriermetrics
           - querierscalar
           - queriercommon
+          - querierai
           - rawexportdata
           - role
           - rootuser
           - serviceaccount
           - querier_json_body
-          - promqlparity
           - querier_skip_resource_fingerprint
           - ttl
         sqlstore-provider:

docs/api/openapi.yml

@@ -8565,6 +8565,7 @@ components:
     TelemetrytypesSource:
       enum:
       - meter
+      - ai
       type: string
     TelemetrytypesTelemetryFieldKey:
       properties:

frontend/src/api/generated/services/sigNoz.schemas.ts

@@ -3631,6 +3631,7 @@ export enum Querybuildertypesv5QueryBuilderQueryGithubComSigNozSignozPkgTypesQue
 }
 export enum TelemetrytypesSourceDTO {
 	meter = 'meter',
+	ai = 'ai',
 }
 export interface Querybuildertypesv5QueryBuilderQueryGithubComSigNozSignozPkgTypesQuerybuildertypesQuerybuildertypesv5LogAggregationDTO {
 	/**

frontend/src/container/InfraMonitoringK8sV2/Base/ColumnHeader.tsx

@@ -51,7 +51,7 @@ function ColumnHeader({
 							<a
 								href={`${DOCS_BASE_URL}${docPath}`}
 								target="_blank"
-								rel="noopener noreferrer"
+								rel="noopener"
 								onClick={(e): void => e.stopPropagation()}
 							>
 								Learn more.

frontend/src/container/InfraMonitoringK8sV2/Base/EntityGroupHeader.tsx

@@ -30,7 +30,7 @@ function EntityGroupHeader({
 							<a
 								href={`${DOCS_BASE_URL}${docPath}`}
 								target="_blank"
-								rel="noopener noreferrer"
+								rel="noopener"
 								onClick={(e): void => e.stopPropagation()}
 							>
 								Learn more.

frontend/src/container/InfraMonitoringK8sV2/Base/K8sBaseDetails.tsx

@@ -63,6 +63,10 @@ import LoadingContainer from '../LoadingContainer';
 
 import '../EntityDetailsUtils/entityDetails.styles.scss';
 import { parseAsString, useQueryState } from 'nuqs';
+import {
+	EntityCountConfig,
+	EntityCountsSection,
+} from './components/EntityCountsSection/EntityCountsSection';
 
 const TimeRangeOffset = 1000000000;
 
@@ -72,6 +76,8 @@ export interface K8sDetailsMetadataConfig<T> {
 	render?: (value: string | number, entity: T) => React.ReactNode;
 }
 
+export type K8sDetailsCountConfig<T> = EntityCountConfig<T>;
+
 export interface K8sDetailsFilters {
 	filter: { expression: string };
 	start: number;
@@ -92,9 +98,12 @@ export interface K8sBaseDetailsProps<T> {
 	getInitialLogTracesExpression: (entity: T) => string;
 	getInitialEventsExpression: (entity: T) => string;
 	metadataConfig: K8sDetailsMetadataConfig<T>[];
+	countsConfig?: K8sDetailsCountConfig<T>[];
+	getCountsFilterExpression?: (entity: T) => string;
 	entityWidgetInfo: {
 		title: string;
 		yAxisUnit: string;
+		docPath?: string;
 	}[];
 	getEntityQueryPayload: (
 		entity: T,
@@ -137,6 +146,8 @@ export default function K8sBaseDetails<T>({
 	getInitialLogTracesExpression,
 	getInitialEventsExpression,
 	metadataConfig,
+	countsConfig,
+	getCountsFilterExpression,
 	entityWidgetInfo,
 	getEntityQueryPayload,
 	queryKeyPrefix,
@@ -479,6 +490,19 @@ export default function K8sBaseDetails<T>({
 								})}
 							</div>
 						</div>
+
+						{countsConfig &&
+							countsConfig.length > 0 &&
+							selectedItem &&
+							getCountsFilterExpression && (
+								<EntityCountsSection
+									entity={entity}
+									countsConfig={countsConfig}
+									selectedItem={selectedItem}
+									filterExpression={getCountsFilterExpression(entity)}
+									closeDrawer={handleClose}
+								/>
+							)}
 					</div>
 
 					{!hideDetailViewTabs && (

frontend/src/container/InfraMonitoringK8sV2/Base/components/EntityCountsSection/EntityCountsSection.module.scss

@@ -0,0 +1,41 @@
+.countsContainer {
+	display: flex;
+	gap: var(--spacing-6);
+	margin-top: var(--spacing-8);
+	flex-wrap: wrap;
+}
+
+.countCard {
+	position: relative;
+	flex: 1;
+	min-width: 120px;
+	max-width: 180px;
+	border: 1px solid var(--l3-border);
+	border-radius: 4px;
+	display: flex;
+	flex-direction: column;
+	gap: var(--spacing-2);
+	padding: var(--spacing-6);
+}
+
+.countLabel {
+	color: var(--l2-foreground);
+	letter-spacing: var(--letter-spacing-wide);
+	text-transform: uppercase;
+}
+
+.countValue {
+	color: var(--l1-foreground);
+	font-family: var(--periscope-font-family-mono);
+	font-size: var(--font-size-xl);
+}
+
+.navigateButton {
+	position: absolute;
+	top: var(--spacing-4);
+	right: var(--spacing-4);
+
+	--button-padding: var(--spacing-1);
+	--button-height: 24px;
+	--button-width: 24px;
+}

frontend/src/container/InfraMonitoringK8sV2/Base/components/EntityCountsSection/EntityCountsSection.tsx

@@ -0,0 +1,107 @@
+import { Tooltip } from 'antd';
+import { Button } from '@signozhq/ui/button';
+import { Typography } from '@signozhq/ui/typography';
+import { Compass } from '@signozhq/icons';
+import { QueryParams } from 'constants/query';
+import { initialQueriesMap } from 'constants/queryBuilder';
+import ROUTES from 'constants/routes';
+import { Link } from 'react-router-dom';
+import { DataSource } from 'types/common/queryBuilder';
+import { v4 as uuid } from 'uuid';
+
+import {
+	INFRA_MONITORING_K8S_PARAMS_KEYS,
+	InfraMonitoringEntity,
+} from '../../../constants';
+import styles from './EntityCountsSection.module.scss';
+
+export interface EntityCountConfig<T> {
+	label: string;
+	getValue: (entity: T) => number;
+	targetCategory: InfraMonitoringEntity;
+}
+
+interface EntityCountsSectionProps<T> {
+	entity: T;
+	countsConfig: EntityCountConfig<T>[];
+	selectedItem: string;
+	filterExpression: string;
+	closeDrawer: () => void;
+}
+
+export function EntityCountsSection<T>({
+	entity,
+	countsConfig,
+	selectedItem,
+	filterExpression,
+	closeDrawer,
+}: EntityCountsSectionProps<T>): JSX.Element {
+	const buildNavigationUrl = (targetCategory: InfraMonitoringEntity): string => {
+		const defaultQuery = initialQueriesMap[DataSource.METRICS];
+
+		const compositeQuery = {
+			...defaultQuery,
+			id: uuid(),
+			builder: {
+				...defaultQuery.builder,
+				queryData: defaultQuery.builder.queryData.map((query) => ({
+					...query,
+					filter: { expression: filterExpression },
+					filters: { items: [], op: 'AND' as const },
+				})),
+			},
+		};
+
+		// TODO(H4ad): After https://github.com/SigNoz/signoz/pull/12038, inherit custom time of drawer to list
+		const urlParams = new URLSearchParams();
+		urlParams.set(INFRA_MONITORING_K8S_PARAMS_KEYS.CATEGORY, targetCategory);
+		urlParams.set(
+			QueryParams.compositeQuery,
+			encodeURIComponent(JSON.stringify(compositeQuery)),
+		);
+
+		return `${ROUTES.INFRASTRUCTURE_MONITORING_KUBERNETES}?${urlParams.toString()}`;
+	};
+
+	return (
+		<div className={styles.countsContainer}>
+			{countsConfig.map((config) => (
+				<div
+					key={config.label}
+					className={styles.countCard}
+					data-testid={`count-card-${config.label.toLowerCase().replace(/\s+/g, '-')}`}
+				>
+					<Typography.Text
+						color="muted"
+						size="small"
+						weight="medium"
+						className={styles.countLabel}
+					>
+						{config.label}
+					</Typography.Text>
+					<Typography.Text className={styles.countValue} size="xl" weight="semibold">
+						{config.getValue(entity) || '-'}
+					</Typography.Text>
+					<Link
+						to={buildNavigationUrl(config.targetCategory)}
+						onClick={closeDrawer}
+						data-testid={`navigate-${config.label.toLowerCase().replace(/\s+/g, '-')}`}
+					>
+						<Tooltip
+							title={`View ${config.label.toLowerCase()} of '${selectedItem}'`}
+							placement="top"
+						>
+							<Button
+								size="icon"
+								variant="ghost"
+								color="secondary"
+								className={styles.navigateButton}
+								prefix={<Compass size={14} />}
+							/>
+						</Tooltip>
+					</Link>
+				</div>
+			))}
+		</div>
+	);
+}

frontend/src/container/InfraMonitoringK8sV2/Clusters/K8sClustersList.tsx

@@ -14,7 +14,9 @@ import { InfraMonitoringEntity } from '../constants';
 import {
 	clusterWidgetInfo,
 	getClusterMetricsQueryPayload,
+	k8sClusterDetailsCountsConfig,
 	k8sClusterDetailsMetadataConfig,
+	k8sClusterGetCountsFilterExpression,
 	k8sClusterGetEntityName,
 	k8sClusterGetSelectedItemExpression,
 	k8sClusterInitialEventsExpression,
@@ -136,6 +138,8 @@ function K8sClustersList({
 				getInitialLogTracesExpression={k8sClusterInitialLogTracesExpression}
 				getInitialEventsExpression={k8sClusterInitialEventsExpression}
 				metadataConfig={k8sClusterDetailsMetadataConfig}
+				countsConfig={k8sClusterDetailsCountsConfig}
+				getCountsFilterExpression={k8sClusterGetCountsFilterExpression}
 				entityWidgetInfo={clusterWidgetInfo}
 				getEntityQueryPayload={getClusterMetricsQueryPayload}
 				queryKeyPrefix="cluster"

frontend/src/container/InfraMonitoringK8sV2/Clusters/constants.ts

@@ -6,9 +6,15 @@ import { EQueryType } from 'types/common/dashboard';
 import { DataSource, ReduceOperators } from 'types/common/queryBuilder';
 import { v4 } from 'uuid';
 
-import { K8sDetailsMetadataConfig } from '../Base/K8sBaseDetails';
+import {
+	K8sDetailsCountConfig,
+	K8sDetailsMetadataConfig,
+} from '../Base/K8sBaseDetails';
 import { formatValueForExpression } from 'components/QueryBuilderV2/utils';
-import { INFRA_MONITORING_ATTR_KEYS } from '../constants';
+import {
+	INFRA_MONITORING_ATTR_KEYS,
+	InfraMonitoringEntity,
+} from '../constants';
 import { SelectedItemParams } from '../hooks';
 import {
 	buildEventsExpression,
@@ -23,6 +29,40 @@ export const k8sClusterGetSelectedItemExpression = (
 export const k8sClusterDetailsMetadataConfig: K8sDetailsMetadataConfig<InframonitoringtypesClusterRecordDTO>[] =
 	[{ label: 'Cluster Name', getValue: (p): string => p.clusterName || '' }];
 
+export const k8sClusterDetailsCountsConfig: K8sDetailsCountConfig<InframonitoringtypesClusterRecordDTO>[] =
+	[
+		{
+			label: 'Namespaces',
+			getValue: (p): number => p.counts?.namespaces ?? 0,
+			targetCategory: InfraMonitoringEntity.NAMESPACES,
+		},
+		{
+			label: 'Nodes',
+			getValue: (p): number => p.counts?.nodes ?? 0,
+			targetCategory: InfraMonitoringEntity.NODES,
+		},
+		{
+			label: 'Deployments',
+			getValue: (p): number => p.counts?.deployments ?? 0,
+			targetCategory: InfraMonitoringEntity.DEPLOYMENTS,
+		},
+		{
+			label: 'StatefulSets',
+			getValue: (p): number => p.counts?.statefulSets ?? 0,
+			targetCategory: InfraMonitoringEntity.STATEFULSETS,
+		},
+		{
+			label: 'DaemonSets',
+			getValue: (p): number => p.counts?.daemonSets ?? 0,
+			targetCategory: InfraMonitoringEntity.DAEMONSETS,
+		},
+		{
+			label: 'Jobs',
+			getValue: (p): number => p.counts?.jobs ?? 0,
+			targetCategory: InfraMonitoringEntity.JOBS,
+		},
+	];
+
 export const k8sClusterInitialEventsExpression = (
 	item: InframonitoringtypesClusterRecordDTO,
 ): string =>
@@ -43,38 +83,54 @@ export const k8sClusterGetEntityName = (
 	item: InframonitoringtypesClusterRecordDTO,
 ): string => item.clusterName || '';
 
+export const k8sClusterGetCountsFilterExpression = (
+	item: InframonitoringtypesClusterRecordDTO,
+): string =>
+	`k8s.cluster.name = ${formatValueForExpression(item.clusterName ?? '')}`;
+
 export const clusterWidgetInfo = [
 	{
 		title: 'CPU Usage, allocatable',
 		yAxisUnit: '',
+		docPath:
+			'/infrastructure-monitoring/kubernetes/clusters/#cpu-usage-allocatable',
 	},
 	{
 		title: 'Memory Usage, allocatable',
 		yAxisUnit: 'bytes',
+		docPath:
+			'/infrastructure-monitoring/kubernetes/clusters/#memory-usage-allocatable',
 	},
 	{
 		title: 'Ready Nodes',
 		yAxisUnit: '',
+		docPath: '/infrastructure-monitoring/kubernetes/clusters/#ready-nodes',
 	},
 	{
 		title: 'NotReady Nodes',
 		yAxisUnit: '',
+		docPath: '/infrastructure-monitoring/kubernetes/clusters/#notready-nodes',
 	},
 	{
 		title: 'Deployments available and desired',
 		yAxisUnit: '',
+		docPath:
+			'/infrastructure-monitoring/kubernetes/clusters/#deployments-available-and-desired',
 	},
 	{
 		title: 'Statefulset pods',
 		yAxisUnit: '',
+		docPath: '/infrastructure-monitoring/kubernetes/clusters/#statefulset-pods',
 	},
 	{
 		title: 'Daemonset nodes',
 		yAxisUnit: '',
+		docPath: '/infrastructure-monitoring/kubernetes/clusters/#daemonset-nodes',
 	},
 	{
 		title: 'Jobs',
 		yAxisUnit: '',
+		docPath: '/infrastructure-monitoring/kubernetes/clusters/#jobs',
 	},
 ];
 

frontend/src/container/InfraMonitoringK8sV2/DaemonSets/constants.ts

@@ -71,18 +71,25 @@ export const daemonSetWidgetInfo = [
 	{
 		title: 'CPU usage, request, limits',
 		yAxisUnit: '',
+		docPath:
+			'/infrastructure-monitoring/kubernetes/daemonsets/#cpu-usage-request-limits',
 	},
 	{
 		title: 'Memory usage, request, limits',
 		yAxisUnit: 'bytes',
+		docPath:
+			'/infrastructure-monitoring/kubernetes/daemonsets/#memory-usage-request-limits',
 	},
 	{
 		title: 'Network IO',
 		yAxisUnit: 'binBps',
+		docPath: '/infrastructure-monitoring/kubernetes/daemonsets/#network-io',
 	},
 	{
 		title: 'Network errors count',
 		yAxisUnit: '',
+		docPath:
+			'/infrastructure-monitoring/kubernetes/daemonsets/#network-errors-count',
 	},
 ];
 

frontend/src/container/InfraMonitoringK8sV2/DaemonSets/table.config.tsx

@@ -140,10 +140,10 @@ export const k8sDaemonSetsColumnsConfig: DaemonSetTableColumnConfig[] = [
 		},
 	},
 	{
-		id: 'node_status',
+		id: 'scheduled_nodes',
 		header: (): React.ReactNode => (
-			<ColumnHeader docPath="/infrastructure-monitoring/kubernetes/daemonsets#node-status">
-				Node Status
+			<ColumnHeader docPath="/infrastructure-monitoring/kubernetes/daemonsets#scheduled-nodes">
+				Scheduled Nodes
 			</ColumnHeader>
 		),
 		accessorFn: (row): number => row.currentNodes,

frontend/src/container/InfraMonitoringK8sV2/Deployments/constants.ts

@@ -71,18 +71,25 @@ export const deploymentWidgetInfo = [
 	{
 		title: 'CPU usage, request, limits',
 		yAxisUnit: '',
+		docPath:
+			'/infrastructure-monitoring/kubernetes/deployments/#cpu-usage-request-limits',
 	},
 	{
 		title: 'Memory usage, request, limits',
 		yAxisUnit: 'bytes',
+		docPath:
+			'/infrastructure-monitoring/kubernetes/deployments/#memory-usage-request-limits',
 	},
 	{
 		title: 'Network IO',
 		yAxisUnit: 'binBps',
+		docPath: '/infrastructure-monitoring/kubernetes/deployments/#network-io',
 	},
 	{
 		title: 'Network error count',
 		yAxisUnit: '',
+		docPath:
+			'/infrastructure-monitoring/kubernetes/deployments/#network-error-count',
 	},
 ];
 

frontend/src/container/InfraMonitoringK8sV2/Deployments/table.config.tsx

@@ -133,10 +133,10 @@ export const k8sDeploymentsColumnsConfig: TableColumnDef<InframonitoringtypesDep
 			},
 		},
 		{
-			id: 'replica_status',
+			id: 'pod_replicas',
 			header: (): React.ReactNode => (
-				<ColumnHeader docPath="/infrastructure-monitoring/kubernetes/deployments#replica-status">
-					Replica Status
+				<ColumnHeader docPath="/infrastructure-monitoring/kubernetes/deployments#pod-replicas">
+					Pod Replicas
 				</ColumnHeader>
 			),
 			accessorFn: (row): number => row.availablePods,

frontend/src/container/InfraMonitoringK8sV2/EntityDetailsUtils/EntityMetrics/ChartHeader.module.scss

@@ -0,0 +1,32 @@
+.chartHeader {
+	display: inline-flex;
+	align-items: center;
+	gap: var(--spacing-2);
+}
+
+.infoIcon {
+	display: inline-flex;
+	align-items: center;
+	color: var(--l2-foreground);
+	cursor: pointer;
+
+	&:hover {
+		color: var(--text-slate-primary);
+	}
+}
+
+.metricsExplorerLink {
+	display: inline-flex;
+	align-items: center;
+	color: var(--l2-foreground);
+	transition: opacity 0.2s;
+
+	&:hover {
+		color: var(--l3-foreground);
+	}
+}
+
+.chartHeaderLabel {
+	font-size: var(--periscope-font-size-base);
+	color: var(--l2-foreground);
+}

frontend/src/container/InfraMonitoringK8sV2/EntityDetailsUtils/EntityMetrics/ChartHeader.tsx

@@ -0,0 +1,83 @@
+import { Link } from 'react-router-dom';
+import { Compass, Info } from '@signozhq/icons';
+import { Tooltip } from 'antd';
+
+import styles from './ChartHeader.module.scss';
+
+const DOCS_BASE_URL = `${process.env.DOCS_BASE_URL}/docs`;
+
+interface ChartHeaderProps {
+	title: string;
+	docPath?: string;
+	tooltip?: string;
+	metricsExplorerUrl?: string;
+	metricsExplorerTestId?: string;
+}
+
+function ChartHeader({
+	title,
+	docPath,
+	tooltip,
+	metricsExplorerUrl,
+	metricsExplorerTestId = 'open-metrics-explorer',
+}: ChartHeaderProps): JSX.Element {
+	const renderInfoIcon = (): React.ReactNode => {
+		if (docPath) {
+			const tooltipTitle = tooltip || 'Not sure what this represents?';
+			return (
+				<Tooltip
+					arrow
+					title={
+						<>
+							{tooltipTitle}{' '}
+							<a
+								href={`${DOCS_BASE_URL}${docPath}`}
+								target="_blank"
+								rel="noopener"
+								onClick={(e): void => e.stopPropagation()}
+							>
+								Learn more.
+							</a>
+						</>
+					}
+				>
+					<span className={styles.infoIcon} data-testid="chart-header-info-icon">
+						<Info size="md" />
+					</span>
+				</Tooltip>
+			);
+		}
+
+		if (tooltip) {
+			return (
+				<Tooltip title={tooltip}>
+					<span className={styles.infoIcon} data-testid="chart-header-info-icon">
+						<Info size="md" />
+					</span>
+				</Tooltip>
+			);
+		}
+
+		return null;
+	};
+
+	return (
+		<div className={styles.chartHeader} data-testid="chart-header">
+			<span className={styles.chartHeaderLabel}>{title}</span>
+			{renderInfoIcon()}
+			{metricsExplorerUrl && (
+				<Tooltip title="Open in Metrics Explorer">
+					<Link
+						to={metricsExplorerUrl}
+						className={styles.metricsExplorerLink}
+						data-testid={metricsExplorerTestId}
+					>
+						<Compass size={14} />
+					</Link>
+				</Tooltip>
+			)}
+		</div>
+	);
+}
+
+export default ChartHeader;

frontend/src/container/InfraMonitoringK8sV2/EntityDetailsUtils/EntityMetrics/EntityMetrics.module.scss

@@ -14,28 +14,6 @@
 	box-sizing: border-box;
 }
 
-.entityMetricsTitleContainer {
-	display: flex;
-	align-items: center;
-	gap: 8px;
-}
-
-.entityMetricsTitle {
-	font-size: var(--periscope-font-size-base);
-	color: var(--l2-foreground);
-}
-
-.metricsExplorerLink {
-	display: flex;
-	align-items: center;
-	color: var(--l2-foreground);
-	transition: opacity 0.2s;
-
-	&:hover {
-		color: var(--l3-foreground);
-	}
-}
-
 .metricsHeader {
 	display: flex;
 	justify-content: flex-end;

frontend/src/container/InfraMonitoringK8sV2/EntityDetailsUtils/EntityMetrics/EntityMetrics.tsx

@@ -1,8 +1,6 @@
 import { useCallback, useMemo, useRef } from 'react';
 import { UseQueryResult } from 'react-query';
-import { Link } from 'react-router-dom';
-import { Compass } from '@signozhq/icons';
-import { Skeleton, Tooltip } from 'antd';
+import { Skeleton } from 'antd';
 import cx from 'classnames';
 import { PANEL_TYPES } from 'constants/queryBuilder';
 import TimeSeries from 'container/DashboardContainer/visualization/charts/TimeSeries/TimeSeries';
@@ -24,6 +22,7 @@ import { MetricRangePayloadProps } from 'types/api/metrics/getQueryRange';
 import { getMetricsExplorerUrl } from 'utils/explorerUtils';
 
 import { buildEntityMetricsChartConfig } from './configBuilder';
+import ChartHeader from './ChartHeader';
 
 import { useEntityMetrics } from './hooks';
 import { isKeyNotFoundError } from '../utils';
@@ -47,6 +46,7 @@ interface EntityMetricsProps<T> {
 	entityWidgetInfo: {
 		title: string;
 		yAxisUnit: string;
+		docPath?: string;
 	}[];
 	getEntityQueryPayload: (
 		node: T,
@@ -207,31 +207,24 @@ function EntityMetrics<T>({
 						key={entityWidgetInfo[idx].title}
 						className={styles.entityMetricsCol}
 					>
-						<div className={styles.entityMetricsTitleContainer}>
-							<span className={styles.entityMetricsTitle}>
-								{entityWidgetInfo[idx].title}
-							</span>
-							{queryPayloads[idx] &&
-								queryPayloads[idx].graphType !== PANEL_TYPES.TABLE && (
-									<Tooltip title="Open in Metrics Explorer">
-										<Link
-											to={getMetricsExplorerUrl({
-												query: queryPayloads[idx].query,
-												...(selectedInterval && selectedInterval !== 'custom'
-													? { relativeTime: selectedInterval }
-													: {
-															startTimeMs: timeRange.startTime * 1000,
-															endTimeMs: timeRange.endTime * 1000,
-														}),
-											})}
-											className={styles.metricsExplorerLink}
-											data-testid={`open-metrics-explorer-${idx}`}
-										>
-											<Compass size={14} />
-										</Link>
-									</Tooltip>
-								)}
-						</div>
+						<ChartHeader
+							title={entityWidgetInfo[idx].title}
+							docPath={entityWidgetInfo[idx].docPath}
+							metricsExplorerUrl={
+								queryPayloads[idx] && queryPayloads[idx].graphType !== PANEL_TYPES.TABLE
+									? getMetricsExplorerUrl({
+											query: queryPayloads[idx].query,
+											...(selectedInterval && selectedInterval !== 'custom'
+												? { relativeTime: selectedInterval }
+												: {
+														startTimeMs: timeRange.startTime * 1000,
+														endTimeMs: timeRange.endTime * 1000,
+													}),
+										})
+									: undefined
+							}
+							metricsExplorerTestId={`open-metrics-explorer-${idx}`}
+						/>
 						<div className={styles.entityMetricsCard} ref={graphRef}>
 							{renderCardContent(query, idx)}
 						</div>

frontend/src/container/InfraMonitoringK8sV2/Jobs/constants.ts

@@ -70,18 +70,22 @@ export const jobWidgetInfo = [
 	{
 		title: 'CPU usage',
 		yAxisUnit: '',
+		docPath: '/infrastructure-monitoring/kubernetes/jobs/#cpu-usage',
 	},
 	{
 		title: 'Memory Usage',
 		yAxisUnit: 'bytes',
+		docPath: '/infrastructure-monitoring/kubernetes/jobs/#memory-usage',
 	},
 	{
 		title: 'Network IO',
 		yAxisUnit: 'binBps',
+		docPath: '/infrastructure-monitoring/kubernetes/jobs/#network-io',
 	},
 	{
 		title: 'Network errors count',
 		yAxisUnit: '',
+		docPath: '/infrastructure-monitoring/kubernetes/jobs/#network-errors-count',
 	},
 ];
 

frontend/src/container/InfraMonitoringK8sV2/Jobs/table.config.tsx

@@ -132,10 +132,10 @@ export const k8sJobsColumnsConfig: JobTableColumnConfig[] = [
 		},
 	},
 	{
-		id: 'completion_status',
+		id: 'completion',
 		header: (): React.ReactNode => (
-			<ColumnHeader docPath="/infrastructure-monitoring/kubernetes/jobs#completion-status">
-				Completion Status
+			<ColumnHeader docPath="/infrastructure-monitoring/kubernetes/jobs#completion">
+				Completions
 			</ColumnHeader>
 		),
 		accessorFn: (row): number => row.successfulPods,
@@ -155,7 +155,7 @@ export const k8sJobsColumnsConfig: JobTableColumnConfig[] = [
 					{
 						value: row.desiredSuccessfulPods,
 						label: 'Desired',
-						color: Color.BG_ROBIN_500,
+						color: Color.BG_AMBER_500,
 					},
 				]}
 			/>

frontend/src/container/InfraMonitoringK8sV2/Namespaces/K8sNamespacesList.tsx

@@ -14,7 +14,9 @@ import { InfraMonitoringEntity } from '../constants';
 import { SelectedItemParams } from '../hooks';
 import {
 	getNamespaceMetricsQueryPayload,
+	k8sNamespaceDetailsCountsConfig,
 	k8sNamespaceDetailsMetadataConfig,
+	k8sNamespaceGetCountsFilterExpression,
 	k8sNamespaceGetEntityName,
 	k8sNamespaceGetSelectedItemExpression,
 	k8sNamespaceInitialEventsExpression,
@@ -137,6 +139,8 @@ function K8sNamespacesList({
 				getInitialLogTracesExpression={k8sNamespaceInitialLogTracesExpression}
 				getInitialEventsExpression={k8sNamespaceInitialEventsExpression}
 				metadataConfig={k8sNamespaceDetailsMetadataConfig}
+				countsConfig={k8sNamespaceDetailsCountsConfig}
+				getCountsFilterExpression={k8sNamespaceGetCountsFilterExpression}
 				entityWidgetInfo={namespaceWidgetInfo}
 				getEntityQueryPayload={getNamespaceMetricsQueryPayload}
 				queryKeyPrefix="namespace"

frontend/src/container/InfraMonitoringK8sV2/Namespaces/constants.ts

@@ -6,9 +6,16 @@ import { EQueryType } from 'types/common/dashboard';
 import { DataSource, ReduceOperators } from 'types/common/queryBuilder';
 import { v4 } from 'uuid';
 
-import { K8sDetailsMetadataConfig } from '../Base/K8sBaseDetails';
-import { INFRA_MONITORING_ATTR_KEYS } from '../constants';
+import {
+	K8sDetailsCountConfig,
+	K8sDetailsMetadataConfig,
+} from '../Base/K8sBaseDetails';
+import {
+	INFRA_MONITORING_ATTR_KEYS,
+	InfraMonitoringEntity,
+} from '../constants';
 import { SelectedItemParams } from '../hooks';
+import { formatValueForExpression } from 'components/QueryBuilderV2/utils';
 import {
 	buildEventsExpression,
 	buildExpressionFromSelectedItemParams,
@@ -33,6 +40,30 @@ export const k8sNamespaceDetailsMetadataConfig: K8sDetailsMetadataConfig<Inframo
 		},
 	];
 
+export const k8sNamespaceDetailsCountsConfig: K8sDetailsCountConfig<InframonitoringtypesNamespaceRecordDTO>[] =
+	[
+		{
+			label: 'Deployments',
+			getValue: (p): number => p.counts?.deployments ?? 0,
+			targetCategory: InfraMonitoringEntity.DEPLOYMENTS,
+		},
+		{
+			label: 'StatefulSets',
+			getValue: (p): number => p.counts?.statefulSets ?? 0,
+			targetCategory: InfraMonitoringEntity.STATEFULSETS,
+		},
+		{
+			label: 'DaemonSets',
+			getValue: (p): number => p.counts?.daemonSets ?? 0,
+			targetCategory: InfraMonitoringEntity.DAEMONSETS,
+		},
+		{
+			label: 'Jobs',
+			getValue: (p): number => p.counts?.jobs ?? 0,
+			targetCategory: InfraMonitoringEntity.JOBS,
+		},
+	];
+
 export const k8sNamespaceInitialEventsExpression = (
 	item: InframonitoringtypesNamespaceRecordDTO,
 ): string =>
@@ -55,46 +86,78 @@ export const k8sNamespaceGetEntityName = (
 	item: InframonitoringtypesNamespaceRecordDTO,
 ): string => item.namespaceName || '';
 
+export const k8sNamespaceGetCountsFilterExpression = (
+	item: InframonitoringtypesNamespaceRecordDTO,
+): string => {
+	const clusterName = item.meta?.[INFRA_MONITORING_ATTR_KEYS.K8S_CLUSTER_NAME];
+	const clauses: string[] = [];
+
+	if (clusterName) {
+		clauses.push(
+			`${INFRA_MONITORING_ATTR_KEYS.K8S_CLUSTER_NAME} = ${formatValueForExpression(clusterName)}`,
+		);
+	}
+	if (item.namespaceName) {
+		clauses.push(
+			`${INFRA_MONITORING_ATTR_KEYS.K8S_NAMESPACE_NAME} = ${formatValueForExpression(item.namespaceName)}`,
+		);
+	}
+
+	return clauses.join(' AND ');
+};
+
 export const namespaceWidgetInfo = [
 	{
 		title: 'CPU Usage (cores)',
 		yAxisUnit: '',
+		docPath: '/infrastructure-monitoring/kubernetes/namespaces/#cpu-usage-cores',
 	},
 	{
 		title: 'Memory Usage (bytes)',
 		yAxisUnit: 'bytes',
+		docPath:
+			'/infrastructure-monitoring/kubernetes/namespaces/#memory-usage-bytes',
 	},
 	{
 		title: 'Pods CPU (top 10)',
 		yAxisUnit: '',
+		docPath: '/infrastructure-monitoring/kubernetes/namespaces/#pods-cpu-top-10',
 	},
 	{
 		title: 'Pods Memory (top 10)',
 		yAxisUnit: 'bytes',
+		docPath:
+			'/infrastructure-monitoring/kubernetes/namespaces/#pods-memory-top-10',
 	},
 	{
 		title: 'Network rate',
 		yAxisUnit: 'binBps',
+		docPath: '/infrastructure-monitoring/kubernetes/namespaces/#network-rate',
 	},
 	{
 		title: 'Network errors',
 		yAxisUnit: '',
+		docPath: '/infrastructure-monitoring/kubernetes/namespaces/#network-errors',
 	},
 	{
-		title: 'StatefulSets',
+		title: 'StatefulSets (pods)',
 		yAxisUnit: '',
+		docPath: '/infrastructure-monitoring/kubernetes/namespaces/#statefulsets',
 	},
 	{
-		title: 'ReplicaSets',
+		title: 'ReplicaSets (pods)',
 		yAxisUnit: '',
+		docPath: '/infrastructure-monitoring/kubernetes/namespaces/#replicasets',
 	},
 	{
-		title: 'DaemonSets',
+		title: 'DaemonSets (nodes)',
 		yAxisUnit: '',
+		docPath: '/infrastructure-monitoring/kubernetes/namespaces/#daemonsets',
 	},
 	{
-		title: 'Deployments',
+		title: 'Deployments (pods)',
 		yAxisUnit: '',
+		docPath: '/infrastructure-monitoring/kubernetes/namespaces/#deployments',
 	},
 ];
 

frontend/src/container/InfraMonitoringK8sV2/Nodes/constants.ts

@@ -57,42 +57,53 @@ export const nodeWidgetInfo = [
 	{
 		title: 'CPU Usage (cores)',
 		yAxisUnit: '',
+		docPath: '/infrastructure-monitoring/kubernetes/nodes/#cpu-usage-cores',
 	},
 	{
 		title: 'Memory Usage (bytes)',
 		yAxisUnit: 'bytes',
+		docPath: '/infrastructure-monitoring/kubernetes/nodes/#memory-usage-bytes',
 	},
 	{
 		title: 'CPU Usage (%)',
 		yAxisUnit: 'percentunit',
+		docPath: '/infrastructure-monitoring/kubernetes/nodes/#cpu-usage-',
 	},
 	{
 		title: 'Memory Usage (%)',
 		yAxisUnit: 'percentunit',
+		docPath: '/infrastructure-monitoring/kubernetes/nodes/#memory-usage-',
 	},
 	{
 		title: 'Pods by CPU (top 10)',
 		yAxisUnit: '',
+		docPath: '/infrastructure-monitoring/kubernetes/nodes/#pods-by-cpu-top-10',
 	},
 	{
 		title: 'Pods by Memory (top 10)',
 		yAxisUnit: 'bytes',
+		docPath: '/infrastructure-monitoring/kubernetes/nodes/#pods-by-memory-top-10',
 	},
 	{
 		title: 'Network error count',
 		yAxisUnit: '',
+		docPath: '/infrastructure-monitoring/kubernetes/nodes/#network-error-count',
 	},
 	{
 		title: 'Network IO rate',
 		yAxisUnit: 'binBps',
+		docPath: '/infrastructure-monitoring/kubernetes/nodes/#network-io-rate',
 	},
 	{
 		title: 'Filesystem usage (bytes)',
 		yAxisUnit: 'bytes',
+		docPath:
+			'/infrastructure-monitoring/kubernetes/nodes/#filesystem-usage-bytes',
 	},
 	{
 		title: 'Filesystem usage (%)',
 		yAxisUnit: 'percentunit',
+		docPath: '/infrastructure-monitoring/kubernetes/nodes/#filesystem-usage-',
 	},
 ];
 

frontend/src/container/InfraMonitoringK8sV2/Pods/constants.ts

@@ -67,54 +67,74 @@ export const podWidgetInfo = [
 	{
 		title: 'CPU Usage (cores)',
 		yAxisUnit: '',
+		docPath: '/infrastructure-monitoring/kubernetes/pods/#cpu-usage-cores-1',
 	},
 	{
 		title: 'CPU Request, Limit Utilization',
 		yAxisUnit: 'percentunit',
+		docPath:
+			'/infrastructure-monitoring/kubernetes/pods/#cpu-request-limit-utilization',
 	},
 	{
 		title: 'Memory Usage (bytes)',
 		yAxisUnit: 'bytes',
+		docPath: '/infrastructure-monitoring/kubernetes/pods/#memory-usage-bytes',
 	},
 	{
 		title: 'Memory Request, Limit Utilization',
 		yAxisUnit: 'percentunit',
+		docPath:
+			'/infrastructure-monitoring/kubernetes/pods/#memory-request-limit-utilization',
 	},
 	{
 		title: 'Memory by State',
 		yAxisUnit: 'bytes',
+		docPath: '/infrastructure-monitoring/kubernetes/pods/#memory-by-state',
 	},
 	{
 		title: 'Memory Major Page Faults',
 		yAxisUnit: '',
+		docPath:
+			'/infrastructure-monitoring/kubernetes/pods/#memory-major-page-faults',
 	},
 	{
 		title: 'CPU Usage by Container (cores)',
 		yAxisUnit: '',
+		docPath:
+			'/infrastructure-monitoring/kubernetes/pods/#cpu-usage-by-container-cores',
 	},
 	{
 		title: 'CPU Request, Limit Utilization by Container',
 		yAxisUnit: 'percentunit',
+		docPath:
+			'/infrastructure-monitoring/kubernetes/pods/#cpu-request-limit-utilization-by-container',
 	},
 	{
 		title: 'Memory Usage by Container (bytes)',
 		yAxisUnit: 'bytes',
+		docPath:
+			'/infrastructure-monitoring/kubernetes/pods/#memory-usage-by-container-bytes',
 	},
 	{
 		title: 'Memory Request, Limit Utilization by Container',
 		yAxisUnit: 'percentunit',
+		docPath:
+			'/infrastructure-monitoring/kubernetes/pods/#memory-request-limit-utilization-by-container',
 	},
 	{
 		title: 'Network rate',
 		yAxisUnit: 'binBps',
+		docPath: '/infrastructure-monitoring/kubernetes/pods/#network-rate',
 	},
 	{
 		title: 'Network errors',
 		yAxisUnit: '',
+		docPath: '/infrastructure-monitoring/kubernetes/pods/#network-errors',
 	},
 	{
 		title: 'File system (bytes)',
 		yAxisUnit: 'bytes',
+		docPath: '/infrastructure-monitoring/kubernetes/pods/#file-system-bytes',
 	},
 ];
 

frontend/src/container/InfraMonitoringK8sV2/Pods/table.config.tsx

@@ -152,7 +152,11 @@ export const k8sPodColumnsConfig: PodTableColumnConfig[] = [
 	},
 	{
 		id: 'podAge',
-		header: 'Age',
+		header: (): React.ReactNode => (
+			<ColumnHeader docPath="/infrastructure-monitoring/kubernetes/pods#age">
+				Age
+			</ColumnHeader>
+		),
 		accessorFn: (row): number => row.podAge,
 		width: { min: 100 },
 		enableSort: false,
@@ -316,7 +320,11 @@ export const k8sPodColumnsConfig: PodTableColumnConfig[] = [
 	},
 	{
 		id: 'namespace',
-		header: 'Namespace',
+		header: (): React.ReactNode => (
+			<ColumnHeader docPath="/infrastructure-monitoring/kubernetes/pods#additional-columns">
+				Namespace
+			</ColumnHeader>
+		),
 		accessorFn: (row): string =>
 			row.meta?.[INFRA_MONITORING_ATTR_KEYS.K8S_NAMESPACE_NAME] || '',
 		width: { default: 100 },
@@ -328,7 +336,11 @@ export const k8sPodColumnsConfig: PodTableColumnConfig[] = [
 	},
 	{
 		id: 'node',
-		header: 'Node',
+		header: (): React.ReactNode => (
+			<ColumnHeader docPath="/infrastructure-monitoring/kubernetes/pods#additional-columns">
+				Node
+			</ColumnHeader>
+		),
 		accessorFn: (row): string =>
 			row.meta?.[INFRA_MONITORING_ATTR_KEYS.K8S_NODE_NAME] || '',
 		width: { default: 100 },
@@ -340,7 +352,11 @@ export const k8sPodColumnsConfig: PodTableColumnConfig[] = [
 	},
 	{
 		id: 'cluster',
-		header: 'Cluster',
+		header: (): React.ReactNode => (
+			<ColumnHeader docPath="/infrastructure-monitoring/kubernetes/pods#additional-columns">
+				Cluster
+			</ColumnHeader>
+		),
 		accessorFn: (row): string =>
 			row.meta?.[INFRA_MONITORING_ATTR_KEYS.K8S_CLUSTER_NAME] || '',
 		width: { default: 100 },

frontend/src/container/InfraMonitoringK8sV2/StatefulSets/constants.ts

@@ -72,26 +72,37 @@ export const statefulSetWidgetInfo = [
 	{
 		title: 'CPU usage, request, limits',
 		yAxisUnit: '',
+		docPath:
+			'/infrastructure-monitoring/kubernetes/statefulsets/#cpu-usage-request-limits',
 	},
 	{
 		title: 'CPU request, limit util (%)',
 		yAxisUnit: 'percentunit',
+		docPath:
+			'/infrastructure-monitoring/kubernetes/statefulsets/#cpu-request-limit-utilization-',
 	},
 	{
 		title: 'Memory usage, request, limits',
 		yAxisUnit: 'bytes',
+		docPath:
+			'/infrastructure-monitoring/kubernetes/statefulsets/#memory-usage-request-limits',
 	},
 	{
 		title: 'Memory request, limit util (%)',
 		yAxisUnit: 'percentunit',
+		docPath:
+			'/infrastructure-monitoring/kubernetes/statefulsets/#memory-request-limit-utilization-',
 	},
 	{
 		title: 'Network IO',
 		yAxisUnit: 'binBps',
+		docPath: '/infrastructure-monitoring/kubernetes/statefulsets/#network-io',
 	},
 	{
 		title: 'Network errors count',
 		yAxisUnit: '',
+		docPath:
+			'/infrastructure-monitoring/kubernetes/statefulsets/#network-errors-count',
 	},
 ];
 

frontend/src/container/InfraMonitoringK8sV2/StatefulSets/table.config.tsx

@@ -141,10 +141,10 @@ export const k8sStatefulSetsColumnsConfig: TableColumnDef<InframonitoringtypesSt
 			},
 		},
 		{
-			id: 'pod_status',
+			id: 'pod_replicas',
 			header: (): React.ReactNode => (
-				<ColumnHeader docPath="/infrastructure-monitoring/kubernetes/statefulsets#pod-status">
-					Pod Status
+				<ColumnHeader docPath="/infrastructure-monitoring/kubernetes/statefulsets#pod-replicas">
+					Pod Replicas
 				</ColumnHeader>
 			),
 			accessorFn: (row): number => row.currentPods,

frontend/src/container/InfraMonitoringK8sV2/Volumes/constants.ts

@@ -69,22 +69,27 @@ export const volumeWidgetInfo = [
 	{
 		title: 'Volume available',
 		yAxisUnit: 'bytes',
+		docPath: '/infrastructure-monitoring/kubernetes/volumes/#volume-available',
 	},
 	{
 		title: 'Volume capacity',
 		yAxisUnit: 'bytes',
+		docPath: '/infrastructure-monitoring/kubernetes/volumes/#volume-capacity',
 	},
 	{
 		title: 'Volume inodes used',
 		yAxisUnit: 'short',
+		docPath: '/infrastructure-monitoring/kubernetes/volumes/#volume-inodes-used',
 	},
 	{
 		title: 'Volume inodes',
 		yAxisUnit: 'short',
+		docPath: '/infrastructure-monitoring/kubernetes/volumes/#volume-inodes',
 	},
 	{
 		title: 'Volume inodes free',
 		yAxisUnit: 'short',
+		docPath: '/infrastructure-monitoring/kubernetes/volumes/#volume-inodes-free',
 	},
 ];
 

frontend/src/container/LogDetailedView/InfraMetrics/constants.ts

@@ -2872,17 +2872,72 @@ export const nodeWidgetInfo = [
 ];
 
 export const hostWidgetInfo = [
-	{ title: 'CPU Usage', yAxisUnit: 'percentunit' },
-	{ title: 'Memory Usage', yAxisUnit: 'bytes' },
-	{ title: 'System Load Average', yAxisUnit: '' },
-	{ title: 'Network usage (bytes)', yAxisUnit: 'bytes' },
-	{ title: 'Network usage (packet/s)', yAxisUnit: 'pps' },
-	{ title: 'Network errors', yAxisUnit: 'short' },
-	{ title: 'Network drops', yAxisUnit: 'short' },
-	{ title: 'Network connections', yAxisUnit: 'short' },
-	{ title: 'System disk io (bytes transferred)', yAxisUnit: 'bytes' },
-	{ title: 'System disk operations/s', yAxisUnit: 'short' },
-	{ title: 'Queue size', yAxisUnit: 'short' },
-	{ title: 'System disk operation time/s', yAxisUnit: 's' },
-	{ title: 'Disk Usage (%) by mountpoint', yAxisUnit: 'percentunit' },
+	{
+		title: 'CPU Usage',
+		yAxisUnit: 'percentunit',
+		docPath: '/infrastructure-monitoring/host-monitoring/#cpu-usage',
+	},
+	{
+		title: 'Memory Usage',
+		yAxisUnit: 'bytes',
+		docPath: '/infrastructure-monitoring/host-monitoring/#memory-usage',
+	},
+	{
+		title: 'System Load Average',
+		yAxisUnit: '',
+		docPath: '/infrastructure-monitoring/host-monitoring/#system-load-average',
+	},
+	{
+		title: 'Network usage (bytes)',
+		yAxisUnit: 'bytes',
+		docPath: '/infrastructure-monitoring/host-monitoring/#network-usage-bytes',
+	},
+	{
+		title: 'Network usage (packet/s)',
+		yAxisUnit: 'pps',
+		docPath: '/infrastructure-monitoring/host-monitoring/#network-usage-packetss',
+	},
+	{
+		title: 'Network errors',
+		yAxisUnit: 'short',
+		docPath: '/infrastructure-monitoring/host-monitoring/#network-errors',
+	},
+	{
+		title: 'Network drops',
+		yAxisUnit: 'short',
+		docPath: '/infrastructure-monitoring/host-monitoring/#network-drops',
+	},
+	{
+		title: 'Network connections',
+		yAxisUnit: 'short',
+		docPath: '/infrastructure-monitoring/host-monitoring/#network-connections',
+	},
+	{
+		title: 'System disk io (bytes transferred)',
+		yAxisUnit: 'bytes',
+		docPath: '/infrastructure-monitoring/host-monitoring/#system-disk-io-bytes',
+	},
+	{
+		title: 'System disk operations/s',
+		yAxisUnit: 'short',
+		docPath:
+			'/infrastructure-monitoring/host-monitoring/#system-disk-operationss',
+	},
+	{
+		title: 'Queue size',
+		yAxisUnit: 'short',
+		docPath: '/infrastructure-monitoring/host-monitoring/#queue-size',
+	},
+	{
+		title: 'System disk operation time/s',
+		yAxisUnit: 's',
+		docPath:
+			'/infrastructure-monitoring/host-monitoring/#system-disk-operation-times',
+	},
+	{
+		title: 'Disk Usage (%) by mountpoint',
+		yAxisUnit: 'percentunit',
+		docPath:
+			'/infrastructure-monitoring/host-monitoring/#disk-usage--by-mountpoint',
+	},
 ];

pkg/flagger/registry.go

@@ -15,8 +15,6 @@ var (
 	FeatureEnableAIObservability  = featuretypes.MustNewName("enable_ai_observability")
 	FeatureEnableMetricsReduction = featuretypes.MustNewName("enable_metrics_reduction")
 	FeatureUseInfraMonitoringV2   = featuretypes.MustNewName("use_infra_monitoring_v2")
-
-	FeatureUsePrometheusClickhouseV2 = featuretypes.MustNewName("use_prometheus_clickhouse_v2")
 )
 
 func MustNewRegistry() featuretypes.Registry {
@@ -117,14 +115,6 @@ func MustNewRegistry() featuretypes.Registry {
 			DefaultVariant: featuretypes.MustNewName("disabled"),
 			Variants:       featuretypes.NewBooleanVariants(),
 		},
-		&featuretypes.Feature{
-			Name:           FeatureUsePrometheusClickhouseV2,
-			Kind:           featuretypes.KindBoolean,
-			Stage:          featuretypes.StageExperimental,
-			Description:    "Runs PromQL queries on the clickhousev2 provider alongside the served engine result and logs any difference; serving is unaffected.",
-			DefaultVariant: featuretypes.MustNewName("disabled"),
-			Variants:       featuretypes.NewBooleanVariants(),
-		},
 	)
 	if err != nil {
 		panic(err)

pkg/modules/llmpricingrule/impllmpricingrule/module.go

@@ -213,18 +213,18 @@ func (module *module) discoverModels(ctx context.Context, orgID valuer.UUID) ([]
 					Spec: qbtypes.QueryBuilderQuery[qbtypes.TraceAggregation]{
 						Name:   "A",
 						Signal: telemetrytypes.SignalTraces,
-						Filter: &qbtypes.Filter{Expression: fmt.Sprintf("%s EXISTS", llmpricingruletypes.GenAIRequestModel)},
+						Filter: &qbtypes.Filter{Expression: fmt.Sprintf("%s EXISTS", telemetrytypes.GenAIRequestModel)},
 						Aggregations: []qbtypes.TraceAggregation{
 							{Expression: "count()", Alias: "spanCount"},
 						},
 						GroupBy: []qbtypes.GroupByKey{
 							{TelemetryFieldKey: telemetrytypes.TelemetryFieldKey{
-								Name:          llmpricingruletypes.GenAIRequestModel,
+								Name:          telemetrytypes.GenAIRequestModel,
 								FieldContext:  telemetrytypes.FieldContextSpan,
 								FieldDataType: telemetrytypes.FieldDataTypeString,
 							}},
 							{TelemetryFieldKey: telemetrytypes.TelemetryFieldKey{
-								Name:          llmpricingruletypes.GenAIProviderName,
+								Name:          telemetrytypes.GenAIProviderName,
 								FieldContext:  telemetrytypes.FieldContextSpan,
 								FieldDataType: telemetrytypes.FieldDataTypeString,
 							}},
@@ -254,9 +254,9 @@ func (module *module) discoverModels(ctx context.Context, orgID valuer.UUID) ([]
 		switch c.Type {
 		case qbtypes.ColumnTypeGroup:
 			switch c.Name {
-			case llmpricingruletypes.GenAIRequestModel:
+			case telemetrytypes.GenAIRequestModel:
 				modelIdx = i
-			case llmpricingruletypes.GenAIProviderName:
+			case telemetrytypes.GenAIProviderName:
 				providerIdx = i
 			}
 		case qbtypes.ColumnTypeAggregation:

pkg/modules/rulestatehistory/implrulestatehistory/condition_builder.go

@@ -9,6 +9,7 @@ import (
 	"github.com/SigNoz/signoz/pkg/querybuilder"
 	qbtypes "github.com/SigNoz/signoz/pkg/types/querybuildertypes/querybuildertypesv5"
 	"github.com/SigNoz/signoz/pkg/types/telemetrytypes"
+	"github.com/SigNoz/signoz/pkg/valuer"
 	"github.com/huandu/go-sqlbuilder"
 )
 
@@ -22,6 +23,7 @@ func newConditionBuilder(fm qbtypes.FieldMapper) qbtypes.ConditionBuilder {
 
 func (c *conditionBuilder) ConditionFor(
 	ctx context.Context,
+	orgID valuer.UUID,
 	startNs uint64,
 	endNs uint64,
 	key *telemetrytypes.TelemetryFieldKey,

pkg/modules/rulestatehistory/implrulestatehistory/field_mapper.go

@@ -8,6 +8,7 @@ import (
 	schema "github.com/SigNoz/signoz-otel-collector/cmd/signozschemamigrator/schema_migrator"
 	qbtypes "github.com/SigNoz/signoz/pkg/types/querybuildertypes/querybuildertypesv5"
 	"github.com/SigNoz/signoz/pkg/types/telemetrytypes"
+	"github.com/SigNoz/signoz/pkg/valuer"
 	"github.com/huandu/go-sqlbuilder"
 )
 
@@ -30,6 +31,12 @@ func newFieldMapper() qbtypes.FieldMapper {
 	return &fieldMapper{}
 }
 
+// CandidateKeys returns nil: rule-state history has no attribute-map fallback, so a
+// context-missing key stays unresolved and the caller errors.
+func (m *fieldMapper) CandidateKeys(_ context.Context, _ valuer.UUID, _ *telemetrytypes.TelemetryFieldKey, _ any, _ map[string][]*telemetrytypes.TelemetryFieldKey) []*telemetrytypes.TelemetryFieldKey {
+	return nil
+}
+
 func (m *fieldMapper) getColumn(_ context.Context, key *telemetrytypes.TelemetryFieldKey) (*schema.Column, error) { //nolint:unparam
 	name := strings.TrimSpace(key.Name)
 	if col, ok := ruleStateHistoryColumns[name]; ok {
@@ -57,7 +64,7 @@ func (m *fieldMapper) ColumnFor(ctx context.Context, _, _ uint64, key *telemetry
 	return []*schema.Column{col}, nil
 }
 
-func (m *fieldMapper) ColumnExpressionFor(ctx context.Context, tsStart, tsEnd uint64, field *telemetrytypes.TelemetryFieldKey, _ map[string][]*telemetrytypes.TelemetryFieldKey) (string, error) {
+func (m *fieldMapper) ColumnExpressionFor(ctx context.Context, _ valuer.UUID, tsStart, tsEnd uint64, field *telemetrytypes.TelemetryFieldKey, _ map[string][]*telemetrytypes.TelemetryFieldKey) (string, error) {
 	colName, err := m.FieldFor(ctx, tsStart, tsEnd, field)
 	if err != nil {
 		return "", err

pkg/prometheus/clickhouseprometheusv2/capture.go

@@ -1,89 +0,0 @@
-package clickhouseprometheusv2
-
-import (
-	"context"
-	"sync"
-
-	"github.com/SigNoz/signoz/pkg/prometheus"
-	"github.com/prometheus/prometheus/model/labels"
-	"github.com/prometheus/prometheus/storage"
-	"github.com/prometheus/prometheus/util/annotations"
-)
-
-// statementRecorder collects the statements a PromQL evaluation would run.
-// Safe for concurrent use: the engine may Select selectors concurrently.
-type statementRecorder struct {
-	mu         sync.Mutex
-	statements []prometheus.CapturedStatement
-}
-
-func (r *statementRecorder) record(query string, args []any) {
-	r.mu.Lock()
-	defer r.mu.Unlock()
-	r.statements = append(r.statements, prometheus.CapturedStatement{Query: query, Args: args})
-}
-
-func (r *statementRecorder) Statements() []prometheus.CapturedStatement {
-	r.mu.Lock()
-	defer r.mu.Unlock()
-	out := make([]prometheus.CapturedStatement, len(r.statements))
-	copy(out, r.statements)
-	return out
-}
-
-type captureQueryable struct {
-	client   *client
-	recorder *statementRecorder
-}
-
-func (c *captureQueryable) Querier(mint, maxt int64) (storage.Querier, error) {
-	return &captureQuerier{
-		querier:  querier{mint: mint, maxt: maxt, client: c.client},
-		recorder: c.recorder,
-	}, nil
-}
-
-// captureQuerier builds the same SQL as the live querier but records it and
-// returns no data. The fingerprint filter always takes the subquery form:
-// without executing the series lookup, the inline literal set is unknown.
-type captureQuerier struct {
-	querier
-	recorder *statementRecorder
-}
-
-func (c *captureQuerier) Select(ctx context.Context, _ bool, hints *storage.SelectHints, matchers ...*labels.Matcher) storage.SeriesSet {
-	if rawQuery, ok := rawSQLQuery(matchers); ok {
-		c.recorder.record(rawQuery, nil)
-		return storage.EmptySeriesSet()
-	}
-
-	start, end := c.window(hints)
-
-	samplesQuery, args, err := buildSamplesQuery(start, end, metricNamesFromMatchers(matchers), nil, matchers, c.lastSamplePerStepFor(ctx, hints))
-	if err != nil {
-		return storage.ErrSeriesSet(err)
-	}
-	c.recorder.record(samplesQuery, args)
-
-	return storage.EmptySeriesSet()
-}
-
-func (c *captureQuerier) LabelValues(context.Context, string, *storage.LabelHints, ...*labels.Matcher) ([]string, annotations.Annotations, error) {
-	return nil, nil, nil
-}
-
-func (c *captureQuerier) LabelNames(context.Context, *storage.LabelHints, ...*labels.Matcher) ([]string, annotations.Annotations, error) {
-	return nil, nil, nil
-}
-
-// metricNamesFromMatchers extracts the statically known metric name, if any.
-// The live path derives names from the matched series; the capture path has
-// no execution results, so only a __name__ equality contributes.
-func metricNamesFromMatchers(matchers []*labels.Matcher) []string {
-	for _, m := range matchers {
-		if m.Name == metricNameLabel && m.Type == labels.MatchEqual && m.Value != "" {
-			return []string{m.Value}
-		}
-	}
-	return nil
-}

pkg/prometheus/clickhouseprometheusv2/client.go

@@ -1,282 +0,0 @@
-package clickhouseprometheusv2
-
-import (
-	"context"
-	"database/sql"
-	"encoding/json"
-	"fmt"
-	"log/slog"
-	"math"
-	"slices"
-
-	"github.com/SigNoz/signoz/pkg/errors"
-	"github.com/SigNoz/signoz/pkg/factory"
-	"github.com/SigNoz/signoz/pkg/prometheus"
-	"github.com/SigNoz/signoz/pkg/telemetrystore"
-	"github.com/SigNoz/signoz/pkg/types/ctxtypes"
-	"github.com/SigNoz/signoz/pkg/types/instrumentationtypes"
-	"github.com/SigNoz/signoz/pkg/types/telemetrytypes"
-	"github.com/prometheus/prometheus/model/labels"
-	promValue "github.com/prometheus/prometheus/model/value"
-)
-
-// seriesLookup is a series-lookup result: matched fingerprints with their
-// labels, and the distinct metric names seen on them.
-type seriesLookup struct {
-	fingerprints map[uint64]labels.Labels
-	metricNames  []string
-}
-
-// client executes the series, samples and raw queries against ClickHouse.
-type client struct {
-	settings       factory.ScopedProviderSettings
-	telemetryStore telemetrystore.TelemetryStore
-	cfg            prometheus.ClickhouseV2Config
-	lookbackMs     int64
-}
-
-func newClient(settings factory.ScopedProviderSettings, telemetryStore telemetrystore.TelemetryStore, cfg prometheus.Config) *client {
-	lookback := cfg.LookbackDelta
-	if lookback <= 0 {
-		// Mirror the engine: promql defaults an unset lookback to 5m.
-		lookback = defaultLookbackDelta
-	}
-	return &client{
-		settings:       settings,
-		telemetryStore: telemetryStore,
-		cfg:            cfg.ClickhouseV2,
-		lookbackMs:     lookback.Milliseconds(),
-	}
-}
-
-func (c *client) withContext(ctx context.Context, functionName string) context.Context {
-	return ctxtypes.NewContextWithCommentVals(ctx, map[string]string{
-		instrumentationtypes.TelemetrySignal:  telemetrytypes.SignalMetrics.StringValue(),
-		instrumentationtypes.CodeNamespace:    "clickhouse-prometheus-v2",
-		instrumentationtypes.CodeFunctionName: functionName,
-	})
-}
-
-// selectSeries runs the series lookup for the given matchers and window.
-func (c *client) selectSeries(ctx context.Context, query string, args []any) (*seriesLookup, error) {
-	ctx = c.withContext(ctx, "selectSeries")
-	rows, err := c.telemetryStore.ClickhouseDB().Query(ctx, query, args...)
-	if err != nil {
-		return nil, err
-	}
-	defer rows.Close()
-
-	lookup := &seriesLookup{fingerprints: make(map[uint64]labels.Labels)}
-	names := make(map[string]struct{})
-
-	var fingerprint uint64
-	var labelsJSON string
-	for rows.Next() {
-		if err := rows.Scan(&fingerprint, &labelsJSON); err != nil {
-			return nil, err
-		}
-		lset, err := unmarshalLabels(labelsJSON)
-		if err != nil {
-			return nil, err
-		}
-		lookup.fingerprints[fingerprint] = lset
-		if name := lset.Get(metricNameLabel); name != "" {
-			names[name] = struct{}{}
-		}
-		if c.cfg.MaxFetchedSeries > 0 && len(lookup.fingerprints) > c.cfg.MaxFetchedSeries {
-			return nil, errors.NewInvalidInputf(
-				errors.CodeInvalidInput,
-				"promql selector matched more than %d series; narrow the label matchers or raise prometheus::clickhousev2::max_fetched_series",
-				c.cfg.MaxFetchedSeries,
-			)
-		}
-	}
-	if err := rows.Err(); err != nil {
-		return nil, err
-	}
-
-	for name := range names {
-		lookup.metricNames = append(lookup.metricNames, name)
-	}
-	slices.Sort(lookup.metricNames)
-
-	return lookup, nil
-}
-
-// unmarshalLabels parses the labels JSON column. Unlike v1, the fingerprint
-// is not injected as a synthetic label (it would take part in `without (...)`
-// grouping and vector matching) and empty-valued labels are dropped: an empty
-// label value means "label absent" in Prometheus, and upstream never produces
-// such labels, but stored attribute JSON can carry them.
-func unmarshalLabels(s string) (labels.Labels, error) {
-	m := make(map[string]string)
-	if err := json.Unmarshal([]byte(s), &m); err != nil {
-		return labels.EmptyLabels(), err
-	}
-	builder := labels.NewScratchBuilder(len(m))
-	for k, v := range m {
-		if v == "" {
-			continue
-		}
-		builder.Add(k, v)
-	}
-	builder.Sort()
-	return builder.Labels(), nil
-}
-
-// selectSamples executes a samples query (raw or last-sample-per-step; both
-// produce the same column shape) and assembles the per-series sample slices.
-// Rows arrive ordered by (fingerprint, unix_milli). Rows whose fingerprint
-// is missing from the lookup are skipped (possible in the subquery filter
-// mode, where the fingerprint filter re-runs after the lookup and can see
-// series born in between). Stale flags map to the engine's StaleNaN.
-// Duplicate timestamps pass through as stored: upstream Prometheus cannot
-// produce them (its TSDB rejects them at ingest), our ingest can under
-// at-least-once retries, and v1 feeds them to the engine as-is —
-// deduplicating here would make this provider silently disagree with both
-// v1 and the transpiled statements over the same dirty data. Uniqueness
-// belongs to the ingest layer.
-func (c *client) selectSamples(ctx context.Context, query string, args []any, lookup *seriesLookup) ([]*series, error) {
-	ctx = c.withContext(ctx, "selectSamples")
-	rows, err := c.telemetryStore.ClickhouseDB().Query(ctx, query, args...)
-	if err != nil {
-		return nil, err
-	}
-	defer rows.Close()
-
-	var (
-		result       []*series
-		current      *series
-		fingerprint  uint64
-		prevFp       uint64
-		timestampMs  int64
-		val          float64
-		flags        uint32
-		first        = true
-		haveCurrent  bool
-		staleMarker  = math.Float64frombits(promValue.StaleNaN)
-		maxSamples   = c.cfg.MaxFetchedSamples
-		fetched      int64
-		unknownCount int
-	)
-
-	for rows.Next() {
-		if err := rows.Scan(&fingerprint, &timestampMs, &val, &flags); err != nil {
-			return nil, err
-		}
-
-		fetched++
-		if maxSamples > 0 && fetched > maxSamples {
-			return nil, errors.NewInvalidInputf(
-				errors.CodeInvalidInput,
-				"promql query would fetch more than %d samples; narrow the selector or time range, or raise prometheus::clickhousev2::max_fetched_samples",
-				maxSamples,
-			)
-		}
-
-		if first || fingerprint != prevFp {
-			first = false
-			prevFp = fingerprint
-			lset, ok := lookup.fingerprints[fingerprint]
-			if !ok {
-				unknownCount++
-				haveCurrent = false
-				continue
-			}
-			current = &series{lset: lset}
-			result = append(result, current)
-			haveCurrent = true
-		}
-		if !haveCurrent {
-			// Remaining rows of a fingerprint missing from the lookup.
-			continue
-		}
-
-		if flags&1 == 1 {
-			val = staleMarker
-		}
-		current.ts = append(current.ts, timestampMs)
-		current.vs = append(current.vs, val)
-	}
-	if err := rows.Err(); err != nil {
-		return nil, err
-	}
-
-	if unknownCount > 0 {
-		c.settings.Logger().DebugContext(ctx, "skipped samples of fingerprints missing from series lookup",
-			slog.Int("unknown_fingerprints", unknownCount))
-	}
-
-	return result, nil
-}
-
-// queryRaw supports the {job="rawsql", query="..."} escape hatch: the value of
-// the query matcher runs as-is, each row becoming a single-sample series
-// stamped at the query end. Column "value" is the sample value; every other
-// column is a label.
-func (c *client) queryRaw(ctx context.Context, query string, ts int64) ([]*series, error) {
-	ctx = c.withContext(ctx, "queryRaw")
-	rows, err := c.telemetryStore.ClickhouseDB().Query(ctx, query)
-	if err != nil {
-		return nil, err
-	}
-	defer rows.Close()
-
-	columns := rows.Columns()
-	targets := make([]any, len(columns))
-	for i := range targets {
-		targets[i] = new(scanner)
-	}
-
-	var result []*series
-	for rows.Next() {
-		if err := rows.Scan(targets...); err != nil {
-			return nil, err
-		}
-		builder := labels.NewScratchBuilder(len(columns))
-		var val float64
-		for i, col := range columns {
-			v := targets[i].(*scanner)
-			if col == "value" {
-				val = v.f
-				continue
-			}
-			builder.Add(col, v.s)
-		}
-		builder.Sort()
-		result = append(result, &series{
-			lset: builder.Labels(),
-			ts:   []int64{ts},
-			vs:   []float64{val},
-		})
-	}
-	if err := rows.Err(); err != nil {
-		return nil, err
-	}
-	return result, nil
-}
-
-var _ sql.Scanner = (*scanner)(nil)
-
-type scanner struct {
-	f float64
-	s string
-}
-
-func (s *scanner) Scan(val any) error {
-	s.f = 0
-	s.s = ""
-
-	s.s = fmt.Sprintf("%v", val)
-	switch val := val.(type) {
-	case int64:
-		s.f = float64(val)
-	case uint64:
-		s.f = float64(val)
-	case float64:
-		s.f = val
-	case []byte:
-		s.s = string(val)
-	}
-	return nil
-}

pkg/prometheus/clickhouseprometheusv2/doc.go

@@ -1,334 +0,0 @@
-// Package clickhouseprometheusv2 is the second-generation ClickHouse-backed
-// Prometheus provider. It exists because the v1 provider fetches every raw
-// sample of a query's union window through the remote-read protobuf layer
-// and hands it to the engine — the cost is a function of ingested data, not
-// of the question asked, which is how a dashboard of PromQL panels takes an
-// instance down.
-//
-// Every query runs in one of two ways, decided per query:
-//
-//   - Transpiled: the query is evaluated entirely inside ClickHouse and only
-//     final (or near-final) per-group grid arrays come back, built on the
-//     timeSeries*ToGrid aggregate functions (the supported ClickHouse floor
-//     is >= 25.6, so they are assumed available).
-//   - Engine: the stock promql engine evaluates over this package's native
-//     storage.Querier. This is the path for everything not transpilable.
-//
-// Correctness is the constraint that shaped both paths: a PromQL result that
-// differs from upstream Prometheus is a lost user, so anything that cannot
-// reproduce engine semantics exactly falls back rather than approximate.
-// The rest of this comment is the PromQL -> SQL story, because that mapping
-// is where correctness is won or lost.
-//
-// # The evaluation model the SQL must reproduce
-//
-// A PromQL range query is an instant query evaluated at every grid point
-// t_i = start + i*step, i = 0..(end-start)/step. At each t_i:
-//
-//   - an instant selector resolves to the latest sample in the left-open
-//     lookback window (t_i - lookback, t_i], and to nothing when that latest
-//     sample is a stale marker — even if older real samples sit inside the
-//     window;
-//   - a range selector [r] collects every sample in (t_i - r, t_i], stale
-//     markers excluded;
-//   - offset d shifts both windows to (t_i - d - w, t_i - d].
-//
-// The transpilation invariant follows from this: every transpiled construct
-// produces, per output series, one array with exactly one slot per grid
-// point — slot i holds the value at t_i, NULL means absent. This is what
-// makes composition correct, not just convenient: the engine evaluates
-// these operators independently per t_i, so any representation that gets
-// every slot right gets the whole query right, and spatial aggregation over
-// arrays is sound because it combines values that belong to the same t_i by
-// construction. Slot index i maps back to t_i = start + i*step at scan time
-// (toMatrix). Everything below is about filling those slots with exactly
-// the numbers the engine would compute — and each equivalence was validated
-// against the vendored engine on live data before its shape entered the
-// allowlist; anything unproven stays on the engine path.
-//
-// # Classification: finding what a statement can answer
-//
-// classify walks the parsed AST looking for "core units" — maximal subtrees
-// of the shape
-//
-//	[agg by/without (...)] [fn(] selector[range] [offset d] [)] [op scalar]...
-//
-// classifyCore peels that chain from the outside in: an optional
-// sum/min/max/avg/count aggregation, then one of the allowlisted functions
-// or a bare instant selector, then the selector with its offset; on the way
-// out it accumulates number-literal arithmetic, comparisons (including
-// bool) and unary minus into a scalar-op pipeline. A node qualifies only if
-// its type, arguments and children are in the proven set — an allowlist, so
-// an overlooked construct becomes a fallback instead of a wrong number.
-//
-// Three unit kinds come out of this, each with its own SQL form:
-// unitRange (rate, irate, increase, delta, idelta over a range selector),
-// unitInstant (instant vector selection, bare or comparison-filtered) and
-// unitOverTime (avg/min/max/sum/count/last _over_time).
-//
-// If the entire tree is one unit, the plan is "full": the statement's rows
-// are the query result. Otherwise every maximal unit is cut out and replaced
-// in the expression with a synthetic selector __signoz_transpiled_N__, and
-// the rewritten expression runs in the engine over the units' materialized
-// results ("hybrid") — histogram_quantile, topk, or/and/unless and vector
-// matching keep exact engine semantics while their expensive inputs were
-// aggregated server-side.
-//
-// Classification refuses when exact semantics cannot be guaranteed
-// server-side: the @ modifier anywhere and default-resolution subqueries
-// (their resolution is a server runtime setting the transpiler cannot see);
-// steps or ranges that are not whole seconds (the grid functions take
-// whole-second parameters); grouping by or matching on __name__ in hybrid
-// plans (the synthetic name would leak into results); name-keeping units —
-// bare/comparison instant selectors and last_over_time keep their real
-// __name__ (keepsName), which substitution would replace, so they transpile
-// only as full plans; and every function outside the allowlist (changes,
-// resets, quantile_over_time, absent, native-histogram functions, ...).
-//
-// Units inside a fixed-resolution subquery evaluate on the subquery's own
-// grid instead of the query grid: epoch-aligned multiples of the resolution
-// strictly after outerStart - offset - range, ending at outer end - offset —
-// the exact derivation the engine uses, because a grid shifted by one step
-// changes which samples every window sees.
-//
-// # From one unit to one statement
-//
-// buildUnitSQL renders each unit as a single statement. For
-// sum by (pod) (rate(m{job="api"}[5m])) the skeleton is:
-//
-//	SELECT gkey, sumForEach(grid) AS grid FROM (
-//	    SELECT series.gkey AS gkey,
-//	           timeSeriesRateToGrid(<start>, <end>, <step>, <range>)(fromUnixTimestamp64Milli(unix_milli), value) AS grid
-//	    FROM signoz_metrics.distributed_samples_v4 AS points
-//	    INNER JOIN (
-//	        SELECT fingerprint, <group key expr> AS gkey
-//	        FROM signoz_metrics.time_series_v4
-//	        WHERE <series predicates>
-//	        GROUP BY fingerprint, gkey
-//	    ) AS series ON points.fingerprint = series.fingerprint
-//	    WHERE metric_name = ? AND temporality IN ['Cumulative', 'Unspecified']
-//	      AND points.fingerprint IN (<matched fingerprints>)
-//	      AND unix_milli > <start - range> AND unix_milli <= <end>
-//	      AND bitAnd(flags, 1) = 0
-//	    GROUP BY points.fingerprint, series.gkey
-//	) GROUP BY gkey
-//	SETTINGS allow_experimental_ts_to_grid_aggregate_function = 1
-//
-// Reading it inside out:
-//
-// The time window is the selector's semantics verbatim: strict > on the
-// lower bound and <= on the upper is the left-open (t - w, t] rule, with the
-// whole window shifted by the offset. bitAnd(flags, 1) = 0 drops stale
-// markers, which PromQL excludes from range vectors.
-//
-// The inner GROUP BY computes one grid array per series.
-// timeSeriesRateToGrid(start, end, step, range) is a parametric aggregate:
-// fed (timestamp, value) pairs it produces Array(Nullable(Float64)) with one
-// slot per grid point. Correct because it implements the engine's
-// extrapolatedRate decision for decision — counter resets, the zero-point
-// clamp, the extrapolation thresholds, the >= 2 samples rule, the left-open
-// window — verified by feeding identical samples to both and comparing
-// slot for slot: the only difference ever observed is the last bit
-// (ClickHouse's C++ and Go round the same formula differently), which is
-// the floating-point floor, not a semantic gap. irate/delta/idelta map to
-// their own timeSeries*ToGrid functions with the same verification;
-// increase has no function of its own and is emitted as
-// arrayMap(x -> x * <range seconds>, <rate expr>), exact by definition —
-// extrapolatedRate computes the same extrapolated delta for both and
-// divides by the range only when isRate, so multiplying it back is the
-// identity, not an approximation. The grid parameters are rendered as
-// literals, not bound args — they are aggregate-function parameters — and
-// the experimental gate rides as a SETTINGS clause on the statement itself
-// so telemetrystore hooks cannot clobber it.
-//
-// The join annotates each series with its group key: toJSONString of the
-// sorted [label, value] pairs the unit projects, extracted from the stored
-// labels JSON. by keeps the listed labels, without excludes them plus
-// __name__, no aggregation keeps everything minus __name__ unless the unit
-// keeps its name — the engine's name-dropping rules. Correct as a grouping
-// key because the pairs are sorted and empty values are filtered: key
-// equality is then exactly label-set equality on the projection —
-// Prometheus treats an empty label value as the label being absent, and
-// stored attribute JSON can carry empties that must not split groups — and
-// the same canonical string parses back into the output label set
-// (labelsFromGroupKey).
-//
-// The outer GROUP BY is the spatial aggregation: sum/min/max/avg/count
-// by/without become the -ForEach combinators. Element-wise aggregation over
-// grid arrays is the engine's per-t_i aggregation, because slot i of every
-// input array refers to the same t_i; the combinators skip NULLs, which is
-// the engine aggregating only the series present at t_i, and an index where
-// every series is absent stays NULL. Two edges need explicit handling:
-// countForEach wraps in a mapping of 0 back to NULL, because a count over
-// an all-absent index is an absent point, not 0; and a unit without
-// aggregation still passes through maxForEach — the identity for the common
-// one-fingerprint group, and a deterministic NULL-skipping merge when a
-// regex __name__ selector collapses distinct metrics onto one projected
-// label set. One caveat is inherent: summation order over series differs
-// from the engine's, so spatial aggregates can differ in the last ULP —
-// float addition is not associative; no ordering reproduces the engine's
-// bit-exactly from inside a GROUP BY.
-//
-// # Instant selectors: staleness needs two aggregates
-//
-// unitInstant uses window = lookback and must reproduce the shadowing rule:
-// the point is absent when the latest in-window sample is a stale marker.
-// timeSeriesLastToGrid alone cannot express that — skipping stale rows in
-// WHERE would resurrect the older real sample the marker was written to
-// bury. So stale rows stay in the scan for this kind only, and the grid
-// expression compares three aggregates per slot:
-//
-//	arrayMap((tall, tok, vok) -> if(tall IS NULL OR tok IS NULL OR tall != tok, NULL, vok),
-//	         timeSeriesLastToGrid(...)(ts, toFloat64(unix_milli)),                           -- last sample overall
-//	         timeSeriesLastToGridIf(...)(ts, toFloat64(unix_milli), bitAnd(flags, 1) = 0),   -- last non-stale, its timestamp
-//	         timeSeriesLastToGridIf(...)(ts, value, bitAnd(flags, 1) = 0))                   -- last non-stale, its value
-//
-// Correct by cases on a slot's window. No samples at all: both timestamp
-// aggregates are NULL, the slot is NULL — absent, as the engine says. Latest
-// sample non-stale: it is the latest overall and the latest non-stale, the
-// timestamps agree, the slot takes its value — the engine's pick. Latest
-// sample stale: the last-overall timestamp is the marker's, the
-// last-non-stale timestamp is older (or NULL when only markers are in
-// window), they disagree, the slot is NULL — the marker shadows, exactly
-// the engine's rule. Timestamps are unique per series (ingest dedups), so
-// timestamp equality identifies "the same sample" without ambiguity. The
-// -If combinator's applicability to these experimental aggregates was
-// probed before being trusted, not assumed.
-//
-// # Windowed *_over_time: fan-out instead of a grid function
-//
-// avg/min/max/sum/count _over_time aggregate every raw sample in the window,
-// and no timeSeries*ToGrid function computes them. (last_over_time is the
-// exception: the last sample of a range vector — stale markers excluded from
-// range vectors by PromQL, excluded here in WHERE — is exactly
-// timeSeriesLastToGrid.) Instead, each sample is fanned out to every grid
-// index whose window contains it:
-//
-//	ARRAY JOIN range(toUInt64(greatest(0, intDiv(unix_milli - <start> + <step> - 1, <step>))),
-//	                 toUInt64(least(<lastIdx>, intDiv(unix_milli + <range> - 1 - <start>, <step>)) + 1)) AS k
-//
-// Correct because the bounds solve the window condition for k. A sample at
-// ts contributes to slot k iff t_k - range < ts <= t_k. The right side
-// gives t_k >= ts, so the first index is ceil((ts - start)/step) — a sample
-// at exactly t_k belongs to k, the window is right-closed. The left side
-// gives t_k < ts + range, and with millisecond-integer timestamps that is
-// t_k <= ts + range - 1, so the last index is
-// floor((ts + range - 1 - start)/step) — a sample at exactly t_k - range is
-// excluded, the window is left-open. Clamped to the grid, the fan-out
-// therefore lands each sample in exactly the slots whose windows contain
-// it, and GROUP BY (fingerprint, k) with the plain aggregate (avg(value),
-// min(value), ...) computes per slot over precisely the engine's sample
-// multiset — the same numbers, since avg/min/max/sum/count are
-// order-insensitive on a given multiset (sum/avg up to summation order, the
-// float caveat above). A second level assembles the positional array with
-// groupArray + indexOf, mapping missing indices to NULL — groupArrayInsertAt
-// would coerce NULL defaults to 0, which is a value, not absence. The
-// group-key join happens at the initiator here, over rows already reduced
-// to per-(series, index); see the sharding section for why that costs
-// nothing.
-//
-// # Scalar ops, full plans, hybrid plans
-//
-// The scalar-op pipeline applies in Go to the returned arrays
-// (applyScalarOps), slot by slot: arithmetic operators compute, comparisons
-// filter (the slot keeps the vector-side value or becomes NULL) or return
-// 0/1 under bool. Correct trivially: it is the same float64 operation the
-// engine would apply to the same slot value, in the same operator order the
-// AST dictates — running it in Go instead of another SQL layer changes
-// where, not what.
-//
-// A full plan's arrays map straight to the result matrix. A hybrid plan
-// materializes each unit's arrays as synthetic series under its
-// __signoz_transpiled_N__ name and evaluates the rewritten expression over
-// a storage that serves synthetic names from memory and everything else
-// live. Substitution is sound because a unit's output is a plain instant
-// vector to the engine — same values at same timestamps under a different
-// name, and the name cannot matter: plans that group by or match on
-// __name__ were refused at classification, and name-keeping units are never
-// substituted. One subtlety makes it exact: stale markers are written at
-// absent grid points, because the engine's lookback would otherwise
-// resurrect a point from up to lookback earlier — the marker encodes
-// "absent here" the way the engine itself encodes it. Units evaluate
-// concurrently; each is one series lookup plus one grid statement. A step
-// of 0 is an instant query: a single evaluation at end.
-//
-// # Series lookup
-//
-// Both paths resolve matchers the same way, once per selector
-// (selectSeries): __name__ matchers translate to the metric_name column —
-// all four matcher types; the v1 client silently returned nothing for regex
-// metric names — and every other matcher to a JSONExtractString condition on
-// the labels column (applySeriesConditions). Regexes are anchored before
-// they reach match(): PromQL matchers match the whole value, ClickHouse
-// match() searches for a substring, and without anchoring =~"api" would
-// also select "x-api-y". An equality matcher against "" matches series
-// without the label, mirroring PromQL, because JSONExtractString returns ""
-// for missing keys. The series tables hold one row per (fingerprint, bucket)
-// at 1h/6h/1d/1w granularities; timeSeriesTableFor picks the table whose
-// bucket fits the window and rounds the window start down to the bucket
-// boundary. The resulting label sets drop what v1 leaked into results: the
-// synthetic fingerprint label (it would take part in without() grouping and
-// vector matching) and empty-valued labels. MaxFetchedSeries fails the
-// lookup with a typed invalid-input error past the ceiling — v1's behavior
-// for an oversized selector was to buffer everything and OOM, and a 4xx the
-// user can narrow beats a dead process serving nobody.
-//
-// # The engine path
-//
-// Queries that do not transpile run in the stock engine over this package's
-// storage.Querier, which is still not the v1 path. Samples are fetched per
-// selector using the engine's per-selector hints, not the query-wide union
-// window, so foo / foo offset 1d reads two narrow windows instead of the
-// widest one twice. Instant selectors of subquery-free queries fetch only
-// the last sample per step bucket (lastSamplePerStep): buckets anchor at the
-// selector's first evaluation timestamp — recovered from the hints as
-// hints.Start + lookback - 1ms, the inverse of how the engine derives
-// hints.Start — so bucket boundaries coincide with evaluation timestamps and
-// a non-final sample of a bucket can never be the latest sample in
-// (t - lookback, t] for any grid t. Real timestamps are preserved, so the
-// engine's own lookback and staleness handling stay exact. Range selectors
-// always fetch raw — every sample feeds the range function — and the
-// subquery-free proof travels in the context as prometheus.QueryTraits,
-// because subquery selectors evaluate at the subquery's step while the
-// hints carry the top-level step. Row assembly counts rows against
-// MaxFetchedSamples while scanning, keeps the first of consecutive equal
-// timestamps, maps stale flags to the engine's StaleNaN, and merges series
-// with identical label sets (sortAndMerge) — the engine assumes storages
-// never emit duplicates. A {job="rawsql", query="..."} selector bypasses all
-// of this and runs the query matcher's value verbatim.
-//
-// # Sharding
-//
-// samples_v4 and time_series_v4 (and all their rollups) shard on the same
-// key — cityHash64(env, temporality, metric_name, fingerprint) — so a
-// series' samples and catalog rows live on the same shard. The transpiled
-// statement above exploits that: the distributed samples table at the
-// top-level FROM makes ClickHouse rewrite the whole inner query per shard,
-// where the join against the shard-local series table and the per-series
-// grid aggregation run next to the data; the initiator only merges
-// aggregate states and applies the spatial -ForEach step. Same layout as
-// the telemetrymetrics statement builder. Fingerprint filters follow suit:
-// matched sets inline as sorted literals up to inlineFingerprintsLimit
-// (literals engage the samples primary key; sorting keeps statements
-// deterministic), beyond it the group-key join alone restricts — a
-// semi-join on the same predicates would only rescan the series table —
-// except the windowed *_over_time fan-out, which has no join and keeps a
-// shard-local IN subquery rather than expand every series of the metric.
-// The engine path's over-limit filter is the same shard-local subquery, not
-// a GLOBAL broadcast of the matched set. The temporality filter on every
-// samples statement is a semantic no-op — the matched fingerprints already
-// come from those temporalities — that engages the leading samples
-// primary-key column. Delta-temporality series stay invisible to PromQL
-// here exactly as they are in v1: the rollout gate is parity with v1, and
-// making Delta visible is its own change with its own semantics to design —
-// a Delta stream fed to rate() as-if-cumulative would be wrong, not just
-// new.
-//
-// # Observability
-//
-// Every statement carries a log_comment with
-// code.namespace=clickhouse-prometheus-v2 and code.function.name naming the
-// call site (selectSeries, selectSamples, transpiledUnit, LabelValues,
-// LabelNames), so this provider's work is attributable in system.query_log
-// without guessing from query text.
-package clickhouseprometheusv2

pkg/prometheus/clickhouseprometheusv2/last_sample_per_step_test.go

@@ -1,191 +0,0 @@
-package clickhouseprometheusv2
-
-import (
-	"fmt"
-	"math/rand"
-	"sort"
-	"testing"
-
-	"github.com/stretchr/testify/require"
-)
-
-// The lastSamplePerStep correctness argument, executed: for instant selectors, keeping
-// only the last sample of every step bucket (bucket 0 = (start, firstEval],
-// bucket i = (firstEval+(i-1)·step, firstEval+i·step]) yields exactly the
-// same instant-vector selections as the raw samples, for every evaluation
-// timestamp on the grid. The engine picks the latest sample in
-// (t-lookback, t] per evaluation timestamp t and treats a stale marker as
-// absent; both behaviors are emulated here directly.
-
-type tsample struct {
-	ts    int64
-	value float64
-	stale bool
-}
-
-// engineSelect emulates the engine's instant-selector resolution at
-// evaluation timestamp t over samples ordered by timestamp: the latest sample
-// in (t-lookback, t], absent when none or when it is a stale marker.
-func engineSelect(samples []tsample, t, lookbackMs int64) (tsample, bool) {
-	var picked tsample
-	found := false
-	for _, s := range samples {
-		if s.ts > t-lookbackMs && s.ts <= t {
-			picked = s
-			found = true
-		}
-	}
-	if !found || picked.stale {
-		return tsample{}, false
-	}
-	return picked, true
-}
-
-// lastPerStep emulates the last-sample-per-step samples query: group samples into buckets and
-// keep only the last sample of each (ties keep either; ClickHouse argMax over
-// equal keys is unspecified, so generated timestamps are unique).
-func lastPerStep(samples []tsample, firstEvalMs, stepMs int64) []tsample {
-	last := make(map[int64]tsample)
-	for _, s := range samples {
-		var bucket int64
-		if stepMs > 0 && s.ts > firstEvalMs {
-			bucket = (s.ts-firstEvalMs-1)/stepMs + 1
-		}
-		if cur, ok := last[bucket]; !ok || s.ts > cur.ts {
-			last[bucket] = s
-		}
-	}
-	out := make([]tsample, 0, len(last))
-	for _, s := range last {
-		out = append(out, s)
-	}
-	sort.Slice(out, func(i, j int) bool { return out[i].ts < out[j].ts })
-	return out
-}
-
-func TestLastSamplePerStepEquivalence(t *testing.T) {
-	rng := rand.New(rand.NewSource(42))
-
-	for caseIdx := 0; caseIdx < 2000; caseIdx++ {
-		// Random query shape. Units are milliseconds but kept small so bucket
-		// boundaries are hit often.
-		stepMs := []int64{1, 2, 5, 7, 30, 60}[rng.Intn(6)]
-		lookbackMs := []int64{1, 3, 5, 10, 45}[rng.Intn(5)]
-		queryStart := int64(1000)
-		numSteps := rng.Int63n(20)
-		queryEnd := queryStart + numSteps*stepMs + rng.Int63n(stepMs) // grid may not divide the range
-
-		// Engine-derived selector window for instant selectors:
-		// hints.Start = firstEval - (lookback - 1), hints.End = queryEnd.
-		hintsStart := queryStart - (lookbackMs - 1)
-		hintsEnd := queryEnd
-		firstEval := hintsStart + lookbackMs - 1
-		require.Equal(t, queryStart, firstEval)
-
-		// Random samples inside the fetch window [hints.Start, hints.End],
-		// with unique timestamps and occasional stale markers. The sample
-		// count is capped by the window size: timestamps are unique.
-		windowSize := hintsEnd - hintsStart + 1
-		numSamples := rng.Int63n(40)
-		if numSamples > windowSize {
-			numSamples = windowSize
-		}
-		seen := make(map[int64]bool)
-		var samples []tsample
-		for int64(len(samples)) < numSamples {
-			ts := hintsStart + rng.Int63n(windowSize)
-			if seen[ts] {
-				continue
-			}
-			seen[ts] = true
-			samples = append(samples, tsample{ts: ts, value: rng.Float64(), stale: rng.Intn(8) == 0})
-		}
-		sort.Slice(samples, func(i, j int) bool { return samples[i].ts < samples[j].ts })
-
-		reduced := lastPerStep(samples, firstEval, stepMs)
-
-		desc := fmt.Sprintf("case=%d step=%d lookback=%d start=%d end=%d samples=%d",
-			caseIdx, stepMs, lookbackMs, queryStart, queryEnd, len(samples))
-
-		for evalTs := queryStart; evalTs <= queryEnd; evalTs += stepMs {
-			rawPick, rawOK := engineSelect(samples, evalTs, lookbackMs)
-			reducedPick, reducedOK := engineSelect(reduced, evalTs, lookbackMs)
-
-			require.Equal(t, rawOK, reducedOK, "%s eval=%d presence mismatch", desc, evalTs)
-			if rawOK {
-				require.Equal(t, rawPick, reducedPick, "%s eval=%d sample mismatch", desc, evalTs)
-			}
-		}
-	}
-}
-
-// Instant queries (step 0) evaluate once at firstEval == hints.End; lastSamplePerStep
-// collapses to a single bucket over the whole window.
-func TestLastSamplePerStepEquivalenceInstantQuery(t *testing.T) {
-	rng := rand.New(rand.NewSource(7))
-
-	for caseIdx := 0; caseIdx < 500; caseIdx++ {
-		lookbackMs := []int64{1, 3, 5, 10, 45}[rng.Intn(5)]
-		evalTs := int64(1000)
-		hintsStart := evalTs - (lookbackMs - 1)
-		hintsEnd := evalTs
-		firstEval := hintsStart + lookbackMs - 1
-		require.Equal(t, evalTs, firstEval)
-
-		windowSize := hintsEnd - hintsStart + 1
-		numSamples := rng.Int63n(10)
-		if numSamples > windowSize {
-			numSamples = windowSize
-		}
-		seen := make(map[int64]bool)
-		var samples []tsample
-		for int64(len(samples)) < numSamples {
-			ts := hintsStart + rng.Int63n(windowSize)
-			if seen[ts] {
-				continue
-			}
-			seen[ts] = true
-			samples = append(samples, tsample{ts: ts, value: rng.Float64(), stale: rng.Intn(4) == 0})
-		}
-		sort.Slice(samples, func(i, j int) bool { return samples[i].ts < samples[j].ts })
-
-		reduced := lastPerStep(samples, firstEval, 0)
-		require.LessOrEqual(t, len(reduced), 1, "instant reduction must keep at most one sample")
-
-		rawPick, rawOK := engineSelect(samples, evalTs, lookbackMs)
-		reducedPick, reducedOK := engineSelect(reduced, evalTs, lookbackMs)
-		require.Equal(t, rawOK, reducedOK, "case=%d presence mismatch", caseIdx)
-		if rawOK {
-			require.Equal(t, rawPick, reducedPick, "case=%d sample mismatch", caseIdx)
-		}
-	}
-}
-
-// A stale marker that is the latest sample of its bucket must shadow older
-// samples: the engine sees the marker and reports the series absent, exactly
-// as with raw samples. Pre-filtering stale rows would instead resurrect the
-// older sample.
-func TestLastSamplePerStepKeepsStaleShadowing(t *testing.T) {
-	lookbackMs := int64(10)
-	stepMs := int64(5)
-	queryStart := int64(1000)
-
-	samples := []tsample{
-		{ts: 998, value: 1.0},                // bucket 0
-		{ts: 999, stale: true},               // bucket 0: marker shadows 998
-		{ts: 1003, value: 2.0},               // bucket 1
-		{ts: 1004, stale: true},              // bucket 1: marker shadows 1003
-		{ts: 1008, value: 3.0, stale: false}, // bucket 2
-	}
-	firstEval := queryStart
-	reduced := lastPerStep(samples, firstEval, stepMs)
-
-	for evalTs := queryStart; evalTs <= queryStart+2*stepMs; evalTs += stepMs {
-		rawPick, rawOK := engineSelect(samples, evalTs, lookbackMs)
-		reducedPick, reducedOK := engineSelect(reduced, evalTs, lookbackMs)
-		require.Equal(t, rawOK, reducedOK, "eval=%d", evalTs)
-		if rawOK {
-			require.Equal(t, rawPick, reducedPick, "eval=%d", evalTs)
-		}
-	}
-}

pkg/prometheus/clickhouseprometheusv2/provider.go

@@ -1,84 +0,0 @@
-package clickhouseprometheusv2
-
-import (
-	"context"
-	"time"
-
-	"github.com/SigNoz/signoz/pkg/factory"
-	"github.com/SigNoz/signoz/pkg/prometheus"
-	"github.com/SigNoz/signoz/pkg/telemetrystore"
-	"github.com/prometheus/prometheus/promql"
-	"github.com/prometheus/prometheus/storage"
-)
-
-// Provider ties the package together: its own engine and parser, the
-// ClickHouse client behind the native storage.Querier, and the transpiler
-// executor. See the package documentation for what runs where and why. It is
-// exported as a concrete type — pkg/querier holds it directly for shadow
-// comparison and pinned serving, and an interface with a single
-// implementation would only hide that dependency.
-type Provider struct {
-	settings factory.ScopedProviderSettings
-	engine   *prometheus.Engine
-	parser   prometheus.Parser
-	client   *client
-	executor *executor
-}
-
-var (
-	_ prometheus.Prometheus        = (*Provider)(nil)
-	_ prometheus.StatementCapturer = (*Provider)(nil)
-)
-
-func NewFactory(telemetryStore telemetrystore.TelemetryStore) factory.ProviderFactory[prometheus.Prometheus, prometheus.Config] {
-	return factory.NewProviderFactory(factory.MustNewName("clickhousev2"), func(ctx context.Context, providerSettings factory.ProviderSettings, config prometheus.Config) (prometheus.Prometheus, error) {
-		return New(ctx, providerSettings, config, telemetryStore)
-	})
-}
-
-func New(_ context.Context, providerSettings factory.ProviderSettings, config prometheus.Config, telemetryStore telemetrystore.TelemetryStore) (*Provider, error) {
-	settings := factory.NewScopedProviderSettings(providerSettings, "github.com/SigNoz/signoz/pkg/prometheus/clickhouseprometheusv2")
-
-	engine := prometheus.NewEngine(settings.Logger(), config)
-	parser := prometheus.NewParser()
-	client := newClient(settings, telemetryStore, config)
-
-	return &Provider{
-		settings: settings,
-		engine:   engine,
-		parser:   parser,
-		client:   client,
-		executor: &executor{client: client, engine: engine, parser: parser},
-	}, nil
-}
-
-// TryExecuteRange evaluates transpilable query shapes directly in ClickHouse
-// (see transpiler.go). ok=false means the shape is not transpilable and the
-// caller should evaluate through Engine over Storage instead.
-func (p *Provider) TryExecuteRange(ctx context.Context, query string, start, end time.Time, step time.Duration) (promql.Matrix, bool, error) {
-	return p.executor.TryExecuteRange(ctx, query, start, end, step)
-}
-
-func (p *Provider) Engine() *prometheus.Engine {
-	return p.engine
-}
-
-func (p *Provider) Parser() prometheus.Parser {
-	return p.parser
-}
-
-func (p *Provider) Storage() storage.Queryable {
-	return p
-}
-
-func (p *Provider) Querier(mint, maxt int64) (storage.Querier, error) {
-	return &querier{mint: mint, maxt: maxt, client: p.client}, nil
-}
-
-// CapturingStorage implements prometheus.StatementCapturer: a storage that
-// records each selector's SQL without executing it, for the preview path.
-// A fresh recorder per call keeps concurrent dry-runs isolated.
-func (p *Provider) CapturingStorage() (storage.Queryable, prometheus.StatementRecorder) {
-	recorder := &statementRecorder{}
-	return &captureQueryable{client: p.client, recorder: recorder}, recorder
-}

pkg/prometheus/clickhouseprometheusv2/querier.go

@@ -1,233 +0,0 @@
-package clickhouseprometheusv2
-
-import (
-	"context"
-	"fmt"
-	"slices"
-	"sort"
-	"time"
-
-	"github.com/SigNoz/signoz/pkg/prometheus"
-	"github.com/huandu/go-sqlbuilder"
-	"github.com/prometheus/prometheus/model/labels"
-	"github.com/prometheus/prometheus/storage"
-	"github.com/prometheus/prometheus/util/annotations"
-)
-
-// defaultLookbackDelta mirrors promql's default when the config leaves the
-// lookback unset; the engine and the storage must agree on it for
-// last-sample-per-step bucket anchoring.
-const defaultLookbackDelta = 5 * time.Minute
-
-// querier is a native storage.Querier over ClickHouse. Unlike v1 it does not
-// round-trip through the remote-read protobuf machinery: Select builds SQL
-// directly from the matchers and hints, and the result set is assembled once
-// into compact series.
-type querier struct {
-	mint, maxt int64
-	client     *client
-}
-
-var _ storage.Querier = (*querier)(nil)
-
-func (q *querier) Select(ctx context.Context, sortSeries bool, hints *storage.SelectHints, matchers ...*labels.Matcher) storage.SeriesSet {
-	if rawQuery, ok := rawSQLQuery(matchers); ok {
-		_, end := q.window(hints)
-		list, err := q.client.queryRaw(ctx, rawQuery, end)
-		if err != nil {
-			return storage.ErrSeriesSet(err)
-		}
-		if sortSeries {
-			sort.Slice(list, func(i, j int) bool { return labels.Compare(list[i].lset, list[j].lset) < 0 })
-		}
-		return newSeriesSet(list)
-	}
-
-	start, end := q.window(hints)
-
-	seriesQuery, seriesArgs, err := buildSeriesQuery(start, end, matchers)
-	if err != nil {
-		return storage.ErrSeriesSet(err)
-	}
-	lookup, err := q.client.selectSeries(ctx, seriesQuery, seriesArgs)
-	if err != nil {
-		return storage.ErrSeriesSet(err)
-	}
-	if len(lookup.fingerprints) == 0 {
-		return storage.EmptySeriesSet()
-	}
-
-	list, err := q.fetchSamples(ctx, start, end, matchers, lookup, q.lastSamplePerStepFor(ctx, hints))
-	if err != nil {
-		return storage.ErrSeriesSet(err)
-	}
-
-	// Sorting doubles as duplicate-label-set detection, which the engine
-	// depends on storages never emitting; the cost is on series count, not
-	// samples.
-	list = sortAndMerge(list)
-	return newSeriesSet(list)
-}
-
-// LabelValues returns the values of a label across series matching the
-// matchers within the querier window.
-func (q *querier) LabelValues(ctx context.Context, name string, hints *storage.LabelHints, matchers ...*labels.Matcher) ([]string, annotations.Annotations, error) {
-	sb := sqlbuilder.NewSelectBuilder()
-	if name == metricNameLabel {
-		sb.Select("DISTINCT metric_name AS value")
-	} else {
-		sb.Select(fmt.Sprintf("DISTINCT JSONExtractString(labels, %s) AS value", sb.Var(name)))
-	}
-	adjustedStart, table := timeSeriesTableFor(q.mint, q.maxt)
-	sb.From(fmt.Sprintf("%s.%s", databaseName, table))
-	if err := applySeriesConditions(sb, adjustedStart, q.maxt, matchers); err != nil {
-		return nil, nil, err
-	}
-	sb.Where("value != ''")
-	if hints != nil && hints.Limit > 0 {
-		sb.Limit(hints.Limit)
-	}
-
-	query, args := sb.BuildWithFlavor(sqlbuilder.ClickHouse)
-	values, err := q.selectStrings(ctx, "LabelValues", query, args)
-	if err != nil {
-		return nil, nil, err
-	}
-	slices.Sort(values)
-	return values, nil, nil
-}
-
-// LabelNames returns the label names present on series matching the matchers
-// within the querier window.
-func (q *querier) LabelNames(ctx context.Context, hints *storage.LabelHints, matchers ...*labels.Matcher) ([]string, annotations.Annotations, error) {
-	sb := sqlbuilder.NewSelectBuilder()
-	sb.Select("DISTINCT arrayJoin(JSONExtractKeys(labels)) AS name")
-	adjustedStart, table := timeSeriesTableFor(q.mint, q.maxt)
-	sb.From(fmt.Sprintf("%s.%s", databaseName, table))
-	if err := applySeriesConditions(sb, adjustedStart, q.maxt, matchers); err != nil {
-		return nil, nil, err
-	}
-	if hints != nil && hints.Limit > 0 {
-		sb.Limit(hints.Limit)
-	}
-
-	query, args := sb.BuildWithFlavor(sqlbuilder.ClickHouse)
-	names, err := q.selectStrings(ctx, "LabelNames", query, args)
-	if err != nil {
-		return nil, nil, err
-	}
-	slices.Sort(names)
-	return names, nil, nil
-}
-
-func (q *querier) Close() error {
-	return nil
-}
-
-// window returns the per-selector fetch window. The engine sends per-selector
-// bounds in the hints (already adjusted for offset, @, range and lookback);
-// they are always at least as tight as the querier-level mint/maxt, which
-// span the union of all selectors in the query.
-func (q *querier) window(hints *storage.SelectHints) (int64, int64) {
-	if hints != nil && hints.Start != 0 && hints.End != 0 && hints.Start <= hints.End {
-		return hints.Start, hints.End
-	}
-	return q.mint, q.maxt
-}
-
-// lastSamplePerStepFor decides whether the fetch can keep only the last
-// sample per step bucket, and computes the bucket parameters. Requirements:
-//   - the call site attached QueryTraits proving the query has no subquery
-//     (subquery selectors evaluate at the subquery's own step, but hints
-//     carry the top-level step);
-//   - the selector is an instant selector (hints.Range == 0); range selectors
-//     need every raw sample in the window;
-//   - per-selector hints are present.
-//
-// The engine derives hints.Start for instant selectors as
-// firstEval - (lookback - 1ms), so the first evaluation timestamp is
-// recovered as hints.Start + lookback - 1ms. Bucket boundaries then coincide
-// with evaluation timestamps, which is what makes keeping only the last
-// sample per bucket lossless.
-func (q *querier) lastSamplePerStepFor(ctx context.Context, hints *storage.SelectHints) *lastSamplePerStep {
-	if hints == nil || hints.Range != 0 || hints.Start <= 0 {
-		return nil
-	}
-	traits, ok := prometheus.QueryTraitsFromContext(ctx)
-	if !ok || !traits.SubqueryFree {
-		return nil
-	}
-	firstEval := hints.Start + q.client.lookbackMs - 1
-	if firstEval > hints.End {
-		// Defensive: never anchor a bucket past the window.
-		firstEval = hints.End
-	}
-	return &lastSamplePerStep{firstEvalMs: firstEval, stepMs: hints.Step}
-}
-
-// fetchSamples runs the samples query for the matched series. Small sets
-// inline the fingerprints as sorted uint64 literals — literals engage the
-// samples primary key, and sorting keeps the statement deterministic for
-// logging and tests. Larger sets re-run the series predicates as a
-// shard-local IN subquery instead: inlining hundreds of thousands of
-// literals makes the statement itself the bottleneck, while the subquery is
-// a cheap primary-key scan on each shard's own series table (see
-// localTimeSeriesTable for why that is complete).
-func (q *querier) fetchSamples(ctx context.Context, start, end int64, matchers []*labels.Matcher, lookup *seriesLookup, lastPerStep *lastSamplePerStep) ([]*series, error) {
-	var fingerprints []uint64
-	if len(lookup.fingerprints) <= inlineFingerprintsLimit {
-		fingerprints = make([]uint64, 0, len(lookup.fingerprints))
-		for fp := range lookup.fingerprints {
-			fingerprints = append(fingerprints, fp)
-		}
-		slices.Sort(fingerprints)
-	}
-	query, args, err := buildSamplesQuery(start, end, lookup.metricNames, fingerprints, matchers, lastPerStep)
-	if err != nil {
-		return nil, err
-	}
-	return q.client.selectSamples(ctx, query, args, lookup)
-}
-
-func (q *querier) selectStrings(ctx context.Context, fn, query string, args []any) ([]string, error) {
-	ctx = q.client.withContext(ctx, fn)
-	rows, err := q.client.telemetryStore.ClickhouseDB().Query(ctx, query, args...)
-	if err != nil {
-		return nil, err
-	}
-	defer rows.Close()
-
-	var out []string
-	var v string
-	for rows.Next() {
-		if err := rows.Scan(&v); err != nil {
-			return nil, err
-		}
-		out = append(out, v)
-	}
-	if err := rows.Err(); err != nil {
-		return nil, err
-	}
-	return out, nil
-}
-
-// rawSQLQuery detects the {job="rawsql", query="..."} escape hatch.
-func rawSQLQuery(matchers []*labels.Matcher) (string, bool) {
-	if len(matchers) != 2 {
-		return "", false
-	}
-	var hasJob bool
-	var query string
-	for _, m := range matchers {
-		if m.Type == labels.MatchEqual && m.Name == "job" && m.Value == "rawsql" {
-			hasJob = true
-		}
-		if m.Type == labels.MatchEqual && m.Name == "query" {
-			query = m.Value
-		}
-	}
-	if hasJob && query != "" {
-		return query, true
-	}
-	return "", false
-}

pkg/prometheus/clickhouseprometheusv2/querier_test.go

@@ -1,221 +0,0 @@
-package clickhouseprometheusv2
-
-import (
-	"context"
-	"fmt"
-	"testing"
-
-	"github.com/DATA-DOG/go-sqlmock"
-	cmock "github.com/SigNoz/clickhouse-go-mock"
-	"github.com/stretchr/testify/assert"
-	"github.com/stretchr/testify/require"
-
-	"github.com/SigNoz/signoz/pkg/errors"
-	"github.com/SigNoz/signoz/pkg/factory"
-	"github.com/SigNoz/signoz/pkg/instrumentation/instrumentationtest"
-	"github.com/SigNoz/signoz/pkg/prometheus"
-	"github.com/SigNoz/signoz/pkg/telemetrystore"
-	"github.com/SigNoz/signoz/pkg/telemetrystore/telemetrystoretest"
-	"github.com/prometheus/prometheus/model/labels"
-	"github.com/prometheus/prometheus/storage"
-)
-
-var (
-	seriesCols = []cmock.ColumnType{
-		{Name: "fingerprint", Type: "UInt64"},
-		{Name: "labels", Type: "String"},
-	}
-	samplesCols = []cmock.ColumnType{
-		{Name: "fingerprint", Type: "UInt64"},
-		{Name: "unix_milli", Type: "Int64"},
-		{Name: "value", Type: "Float64"},
-		{Name: "flags", Type: "UInt32"},
-	}
-)
-
-func newTestClient(t *testing.T, cfg prometheus.ClickhouseV2Config) (*client, *telemetrystoretest.Provider) {
-	t.Helper()
-	store := telemetrystoretest.New(telemetrystore.Config{Provider: "clickhouse"}, sqlmock.QueryMatcherRegexp)
-	settings := factory.NewScopedProviderSettings(instrumentationtest.New().ToProviderSettings(), "clickhouseprometheusv2_test")
-	promCfg := prometheus.Config{ClickhouseV2: cfg}
-	return newClient(settings, store, promCfg), store
-}
-
-func testMatchers(t *testing.T) []*labels.Matcher {
-	t.Helper()
-	return []*labels.Matcher{
-		mustMatcher(t, labels.MatchEqual, "__name__", "cpu_usage"),
-		mustMatcher(t, labels.MatchEqual, "job", "api"),
-	}
-}
-
-func TestQuerierSelectRawPath(t *testing.T) {
-	c, store := newTestClient(t, prometheus.ClickhouseV2Config{})
-	q := &querier{mint: 1000, maxt: 2000, client: c}
-
-	store.Mock().ExpectQuery("SELECT fingerprint, any\\(labels\\)").WithArgs("cpu_usage", int64(0), int64(2000), "job", "api").WillReturnRows(cmock.NewRows(seriesCols, [][]any{
-		{uint64(42), `{"__name__":"cpu_usage","job":"api","instance":"a"}`},
-		{uint64(7), `{"__name__":"cpu_usage","job":"api","instance":"b"}`},
-	}))
-	// Inline fingerprints (sorted), raw samples: no traits in ctx -> no
-	// last-sample-per-step reduction.
-	store.Mock().ExpectQuery("SELECT fingerprint, unix_milli, value, flags FROM signoz_metrics.distributed_samples_v4 WHERE metric_name = \\? AND temporality IN \\['Cumulative', 'Unspecified'\\] AND fingerprint IN \\(7, 42\\)").
-		WithArgs("cpu_usage", int64(1000), int64(2000)).
-		WillReturnRows(cmock.NewRows(samplesCols, [][]any{
-			{uint64(7), int64(1100), 1.5, uint32(0)},
-			{uint64(7), int64(1200), 2.5, uint32(0)},
-			{uint64(42), int64(1100), 3.5, uint32(1)}, // stale marker
-		}))
-
-	hints := &storage.SelectHints{Start: 1000, End: 2000, Step: 60_000}
-	set := q.Select(context.Background(), false, hints, testMatchers(t)...)
-
-	var got []*series
-	for set.Next() {
-		got = append(got, set.At().(*series))
-	}
-	require.NoError(t, set.Err())
-	require.Len(t, got, 2)
-
-	// Sorted by labels: instance=a (fp 42) before instance=b (fp 7).
-	assert.Equal(t, "a", got[0].lset.Get("instance"))
-	require.Len(t, got[0].ts, 1)
-	assert.True(t, got[0].vs[0] != got[0].vs[0], "stale marker must be NaN") //nolint:testifylint
-
-	assert.Equal(t, "b", got[1].lset.Get("instance"))
-	assert.Equal(t, []int64{1100, 1200}, got[1].ts)
-	assert.Equal(t, []float64{1.5, 2.5}, got[1].vs)
-
-	// No fingerprint label injected.
-	assert.Empty(t, got[0].lset.Get("fingerprint"))
-}
-
-// Wrong gating silently corrupts range functions (a rate over reduced
-// samples loses points), so the decision logic is pinned here even though
-// the helper is unexported: the integration suite would catch it too, but
-// with far worse failure locality.
-func TestLastSamplePerStepFor(t *testing.T) {
-	c, _ := newTestClient(t, prometheus.ClickhouseV2Config{})
-	q := &querier{mint: 0, maxt: 2000, client: c}
-	traitsCtx := prometheus.NewContextWithQueryTraits(context.Background(), prometheus.QueryTraits{SubqueryFree: true})
-
-	tests := []struct {
-		name  string
-		ctx   context.Context
-		hints *storage.SelectHints
-		want  *lastSamplePerStep
-	}{
-		{"no traits in context stays raw", context.Background(), &storage.SelectHints{Start: 1000, End: 2000, Step: 60_000}, nil},
-		{"subquery in the query stays raw", prometheus.NewContextWithQueryTraits(context.Background(), prometheus.QueryTraits{SubqueryFree: false}), &storage.SelectHints{Start: 1000, End: 2000, Step: 60_000}, nil},
-		{"range selector stays raw", traitsCtx, &storage.SelectHints{Start: 1000, End: 2000, Step: 60_000, Range: 300_000}, nil},
-		{"instant selector reduces, anchored at first eval", traitsCtx, &storage.SelectHints{Start: 1000, End: 2_000_000, Step: 60_000}, &lastSamplePerStep{firstEvalMs: 1000 + c.lookbackMs - 1, stepMs: 60_000}},
-		{"anchor never passes the window end", traitsCtx, &storage.SelectHints{Start: 1000, End: 2000, Step: 60_000}, &lastSamplePerStep{firstEvalMs: 2000, stepMs: 60_000}},
-	}
-	for _, tt := range tests {
-		t.Run(tt.name, func(t *testing.T) {
-			assert.Equal(t, tt.want, q.lastSamplePerStepFor(tt.ctx, tt.hints))
-		})
-	}
-}
-
-func TestQuerierSelectSeriesBudget(t *testing.T) {
-	c, store := newTestClient(t, prometheus.ClickhouseV2Config{MaxFetchedSeries: 1})
-	q := &querier{mint: 1000, maxt: 2000, client: c}
-
-	store.Mock().ExpectQuery("SELECT fingerprint, any\\(labels\\)").WithArgs("cpu_usage", int64(0), int64(2000), "job", "api").WillReturnRows(cmock.NewRows(seriesCols, [][]any{
-		{uint64(1), `{"__name__":"cpu_usage","instance":"a"}`},
-		{uint64(2), `{"__name__":"cpu_usage","instance":"b"}`},
-	}))
-
-	set := q.Select(context.Background(), false, &storage.SelectHints{Start: 1000, End: 2000}, testMatchers(t)...)
-	assert.False(t, set.Next())
-	require.Error(t, set.Err())
-	assert.True(t, errors.Ast(set.Err(), errors.TypeInvalidInput), "budget error must be typed invalid input, got %v", set.Err())
-}
-
-func TestQuerierSelectSamplesBudget(t *testing.T) {
-	c, store := newTestClient(t, prometheus.ClickhouseV2Config{MaxFetchedSamples: 2})
-	q := &querier{mint: 1000, maxt: 2000, client: c}
-
-	store.Mock().ExpectQuery("SELECT fingerprint, any\\(labels\\)").WithArgs("cpu_usage", int64(0), int64(2000)).WillReturnRows(cmock.NewRows(seriesCols, [][]any{
-		{uint64(7), `{"__name__":"cpu_usage"}`},
-	}))
-	store.Mock().ExpectQuery("SELECT fingerprint, unix_milli, value, flags").
-		WithArgs("cpu_usage", int64(1000), int64(2000)).
-		WillReturnRows(cmock.NewRows(samplesCols, [][]any{
-			{uint64(7), int64(1100), 1.0, uint32(0)},
-			{uint64(7), int64(1200), 2.0, uint32(0)},
-			{uint64(7), int64(1300), 3.0, uint32(0)},
-		}))
-
-	set := q.Select(context.Background(), false, &storage.SelectHints{Start: 1000, End: 2000},
-		mustMatcher(t, labels.MatchEqual, "__name__", "cpu_usage"))
-	assert.False(t, set.Next())
-	require.Error(t, set.Err())
-	assert.True(t, errors.Ast(set.Err(), errors.TypeInvalidInput))
-}
-
-func TestQuerierSelectSubqueryFilterOverInlineLimit(t *testing.T) {
-	c, store := newTestClient(t, prometheus.ClickhouseV2Config{})
-	q := &querier{mint: 1000, maxt: 2000, client: c}
-
-	seriesRows := make([][]any, inlineFingerprintsLimit+1)
-	for i := range seriesRows {
-		seriesRows[i] = []any{uint64(i + 1), fmt.Sprintf(`{"__name__":"cpu_usage","instance":"i%d"}`, i)}
-	}
-	store.Mock().ExpectQuery("SELECT fingerprint, any\\(labels\\)").WithArgs("cpu_usage", int64(0), int64(2000), "job", "api").WillReturnRows(cmock.NewRows(seriesCols, seriesRows))
-	// The over-limit samples query embeds the semi-join against the
-	// shard-local series table (fingerprint co-locality), not a GLOBAL
-	// broadcast; args follow placeholder order — samples metric name, then
-	// the semi-join's series predicates, then the samples window bounds.
-	store.Mock().ExpectQuery("fingerprint IN \\(SELECT fingerprint FROM signoz_metrics\\.time_series_v4").
-		WithArgs("cpu_usage", "cpu_usage", int64(0), int64(2000), "job", "api", int64(1000), int64(2000)).
-		WillReturnRows(cmock.NewRows(samplesCols, [][]any{}))
-
-	set := q.Select(context.Background(), false, &storage.SelectHints{Start: 1000, End: 2000}, testMatchers(t)...)
-	assert.False(t, set.Next())
-	require.NoError(t, set.Err())
-}
-
-func TestQuerierSelectRawSQLPassthrough(t *testing.T) {
-	c, store := newTestClient(t, prometheus.ClickhouseV2Config{})
-	q := &querier{mint: 1000, maxt: 2000, client: c}
-
-	rawCols := []cmock.ColumnType{
-		{Name: "le", Type: "String"},
-		{Name: "value", Type: "Float64"},
-	}
-	store.Mock().ExpectQuery("SELECT le, avg\\(v\\) AS value FROM t").WillReturnRows(cmock.NewRows(rawCols, [][]any{
-		{"0.5", 12.5},
-	}))
-
-	set := q.Select(context.Background(), false, &storage.SelectHints{Start: 1000, End: 2000},
-		mustMatcher(t, labels.MatchEqual, "job", "rawsql"),
-		mustMatcher(t, labels.MatchEqual, "query", "SELECT le, avg(v) AS value FROM t"),
-	)
-
-	require.True(t, set.Next())
-	s := set.At()
-	assert.Equal(t, "0.5", s.Labels().Get("le"))
-	it := s.Iterator(nil)
-	require.NotNil(t, it)
-	_, v := func() (int64, float64) { it.Next(); return it.At() }()
-	assert.Equal(t, 12.5, v)
-	assert.False(t, set.Next())
-}
-
-func TestCaptureQuerierRecordsWithoutExecuting(t *testing.T) {
-	c, _ := newTestClient(t, prometheus.ClickhouseV2Config{})
-	recorder := &statementRecorder{}
-	cq := &captureQuerier{querier: querier{mint: 1000, maxt: 2000, client: c}, recorder: recorder}
-
-	ctx := prometheus.NewContextWithQueryTraits(context.Background(), prometheus.QueryTraits{SubqueryFree: true})
-	set := cq.Select(ctx, false, &storage.SelectHints{Start: 1000, End: 2000, Step: 60_000}, testMatchers(t)...)
-	assert.False(t, set.Next())
-	require.NoError(t, set.Err())
-
-	statements := recorder.Statements()
-	require.Len(t, statements, 1)
-	assert.Contains(t, statements[0].Query, "IN (SELECT fingerprint FROM signoz_metrics.time_series_v4")
-	assert.Contains(t, statements[0].Query, "argMax(value, unix_milli)")
-}

pkg/prometheus/clickhouseprometheusv2/seriesset.go

@@ -1,184 +0,0 @@
-package clickhouseprometheusv2
-
-import (
-	"sort"
-
-	"github.com/prometheus/prometheus/model/histogram"
-	"github.com/prometheus/prometheus/model/labels"
-	"github.com/prometheus/prometheus/storage"
-	"github.com/prometheus/prometheus/tsdb/chunkenc"
-	"github.com/prometheus/prometheus/util/annotations"
-)
-
-// series is one time series with samples stored as parallel slices, ordered
-// by timestamp. The compact layout avoids per-sample allocations and keeps
-// iteration cache friendly.
-type series struct {
-	lset labels.Labels
-	ts   []int64
-	vs   []float64
-}
-
-var _ storage.Series = (*series)(nil)
-
-func (s *series) Labels() labels.Labels {
-	return s.lset
-}
-
-func (s *series) Iterator(it chunkenc.Iterator) chunkenc.Iterator {
-	if fit, ok := it.(*floatIterator); ok {
-		fit.reset(s)
-		return fit
-	}
-	fit := &floatIterator{}
-	fit.reset(s)
-	return fit
-}
-
-// floatIterator implements chunkenc.Iterator over a series' sample slices.
-type floatIterator struct {
-	s *series
-	i int
-}
-
-var _ chunkenc.Iterator = (*floatIterator)(nil)
-
-func (it *floatIterator) reset(s *series) {
-	it.s = s
-	it.i = -1
-}
-
-func (it *floatIterator) Next() chunkenc.ValueType {
-	it.i++
-	if it.i >= len(it.s.ts) {
-		return chunkenc.ValNone
-	}
-	return chunkenc.ValFloat
-}
-
-func (it *floatIterator) Seek(t int64) chunkenc.ValueType { //nolint:govet // stdmethods flags io.Seeker; this is chunkenc.Iterator's Seek
-	if it.i < 0 {
-		it.i = 0
-	}
-	if it.i >= len(it.s.ts) {
-		return chunkenc.ValNone
-	}
-	// The current position, once valid, must not move backwards.
-	if it.s.ts[it.i] >= t {
-		return chunkenc.ValFloat
-	}
-	it.i += sort.Search(len(it.s.ts)-it.i, func(j int) bool {
-		return it.s.ts[it.i+j] >= t
-	})
-	if it.i >= len(it.s.ts) {
-		return chunkenc.ValNone
-	}
-	return chunkenc.ValFloat
-}
-
-func (it *floatIterator) At() (int64, float64) {
-	return it.s.ts[it.i], it.s.vs[it.i]
-}
-
-func (it *floatIterator) AtHistogram(*histogram.Histogram) (int64, *histogram.Histogram) {
-	return 0, nil
-}
-
-func (it *floatIterator) AtFloatHistogram(*histogram.FloatHistogram) (int64, *histogram.FloatHistogram) {
-	return 0, nil
-}
-
-func (it *floatIterator) AtT() int64 {
-	return it.s.ts[it.i]
-}
-
-// AtST returns the current start timestamp; not tracked by this storage.
-func (it *floatIterator) AtST() int64 {
-	return 0
-}
-
-func (it *floatIterator) Err() error {
-	return nil
-}
-
-// seriesSet iterates a fully materialized, label-sorted list of series.
-type seriesSet struct {
-	series []*series
-	i      int
-}
-
-var _ storage.SeriesSet = (*seriesSet)(nil)
-
-func newSeriesSet(list []*series) *seriesSet {
-	return &seriesSet{series: list, i: -1}
-}
-
-func (s *seriesSet) Next() bool {
-	s.i++
-	return s.i < len(s.series)
-}
-
-func (s *seriesSet) At() storage.Series {
-	return s.series[s.i]
-}
-
-func (s *seriesSet) Err() error {
-	return nil
-}
-
-func (s *seriesSet) Warnings() annotations.Annotations {
-	return nil
-}
-
-// sortAndMerge orders series by label set and merges series whose label sets
-// are identical. Distinct fingerprints can carry identical label sets (e.g.
-// series differing only in a non-label dimension); Prometheus storages never
-// expose duplicate label sets to the engine, so merge their samples by
-// timestamp, keeping the first sample on ties.
-func sortAndMerge(list []*series) []*series {
-	if len(list) < 2 {
-		return list
-	}
-	sort.Slice(list, func(i, j int) bool {
-		return labels.Compare(list[i].lset, list[j].lset) < 0
-	})
-	out := list[:1]
-	for _, s := range list[1:] {
-		last := out[len(out)-1]
-		if labels.Compare(last.lset, s.lset) != 0 {
-			out = append(out, s)
-			continue
-		}
-		merged := mergeSamples(last, s)
-		out[len(out)-1] = merged
-	}
-	return out
-}
-
-func mergeSamples(a, b *series) *series {
-	ts := make([]int64, 0, len(a.ts)+len(b.ts))
-	vs := make([]float64, 0, len(a.ts)+len(b.ts))
-	i, j := 0, 0
-	for i < len(a.ts) && j < len(b.ts) {
-		switch {
-		case a.ts[i] < b.ts[j]:
-			ts = append(ts, a.ts[i])
-			vs = append(vs, a.vs[i])
-			i++
-		case a.ts[i] > b.ts[j]:
-			ts = append(ts, b.ts[j])
-			vs = append(vs, b.vs[j])
-			j++
-		default:
-			ts = append(ts, a.ts[i])
-			vs = append(vs, a.vs[i])
-			i++
-			j++
-		}
-	}
-	ts = append(ts, a.ts[i:]...)
-	vs = append(vs, a.vs[i:]...)
-	ts = append(ts, b.ts[j:]...)
-	vs = append(vs, b.vs[j:]...)
-	return &series{lset: a.lset, ts: ts, vs: vs}
-}

pkg/prometheus/clickhouseprometheusv2/sql.go

@@ -1,199 +0,0 @@
-package clickhouseprometheusv2
-
-import (
-	"fmt"
-	"strconv"
-	"strings"
-
-	"github.com/SigNoz/signoz/pkg/errors"
-	"github.com/SigNoz/signoz/pkg/query-service/constants"
-	"github.com/huandu/go-sqlbuilder"
-	"github.com/prometheus/prometheus/model/labels"
-)
-
-// inlineFingerprintsLimit is the largest matched-series count inlined into
-// the samples query as literals. Literals engage the samples primary key and
-// avoid a second series-table scan; past a few thousand the statement itself
-// becomes the cost, and the shard-local subquery filter wins. Not
-// configurable: the crossover depends on statement parsing, not on any
-// property of a deployment an operator could know better.
-const inlineFingerprintsLimit = 5_000
-
-// buildSeriesQuery renders the series lookup: one row per matched fingerprint
-// with its labels.
-func buildSeriesQuery(start, end int64, matchers []*labels.Matcher) (string, []any, error) {
-	adjustedStart, table := timeSeriesTableFor(start, end)
-	sb := sqlbuilder.NewSelectBuilder()
-	sb.Select("fingerprint", "any(labels)")
-	sb.From(fmt.Sprintf("%s.%s", databaseName, table))
-	if err := applySeriesConditions(sb, adjustedStart, end, matchers); err != nil {
-		return "", nil, err
-	}
-	sb.GroupBy("fingerprint")
-	query, args := sb.BuildWithFlavor(sqlbuilder.ClickHouse)
-	return query, args, nil
-}
-
-// buildSamplesQuery renders the samples fetch for the series selected by the
-// series lookup. Small matched sets pass inlineFingerprints — sorted uint64
-// literals that engage the samples primary key; nil means the set exceeded
-// the inline limit, and the filter becomes a semi-join re-running the series
-// predicates against the shard-local series table (complete by fingerprint
-// co-locality, see localTimeSeriesTable; a GLOBAL broadcast of the matched
-// set would ship it to every shard instead). metricNames narrows the
-// primary-key scan; when the selector had no __name__ equality, the names
-// observed on the matched series are used. A non-nil lastPerStep groups to
-// one (the last) sample per step bucket.
-func buildSamplesQuery(start, end int64, metricNames []string, inlineFingerprints []uint64, matchers []*labels.Matcher, lastPerStep *lastSamplePerStep) (string, []any, error) {
-	sb := sqlbuilder.NewSelectBuilder()
-	if lastPerStep != nil {
-		// Aliases must not shadow source columns: ClickHouse resolves aliases
-		// in WHERE too, and "max(unix_milli) AS unix_milli" would put an
-		// aggregate into the WHERE clause (error 184).
-		sb.Select("fingerprint", "max(unix_milli) AS ts", "argMax(value, unix_milli) AS val", "argMax(flags, unix_milli) AS fl")
-	} else {
-		sb.Select("fingerprint", "unix_milli", "value", "flags")
-	}
-	sb.From(fmt.Sprintf("%s.%s", databaseName, distributedSamplesV4))
-
-	switch len(metricNames) {
-	case 0:
-		// No name constraint derivable; correct but unable to use the
-		// metric_name primary-key prefix.
-	case 1:
-		sb.Where(sb.EQ("metric_name", metricNames[0]))
-	default:
-		sb.Where(sb.In("metric_name", sqlbuilder.List(metricNames)))
-	}
-	// temporality precedes metric_name in the samples primary key; the
-	// fingerprints already come from these temporalities, so this only helps
-	// granule pruning.
-	sb.Where("temporality IN ['Cumulative', 'Unspecified']")
-	if inlineFingerprints != nil {
-		sb.Where("fingerprint " + inlineFingerprintFilter(inlineFingerprints))
-	} else {
-		sub := sqlbuilder.NewSelectBuilder()
-		sub.Select("fingerprint")
-		adjustedStart, table := timeSeriesTableFor(start, end)
-		sub.From(fmt.Sprintf("%s.%s", databaseName, localTimeSeriesTable(table)))
-		if err := applySeriesConditions(sub, adjustedStart, end, matchers); err != nil {
-			return "", nil, err
-		}
-		sb.Where(sb.In("fingerprint", sub))
-	}
-	sb.Where(sb.GTE("unix_milli", start), sb.LTE("unix_milli", end))
-
-	if lastPerStep != nil {
-		sb.GroupBy("fingerprint")
-		if expr := lastPerStep.bucketExpr(); expr != "" {
-			sb.GroupBy(expr)
-		}
-		sb.OrderBy("fingerprint", "ts")
-	} else {
-		sb.OrderBy("fingerprint", "unix_milli")
-	}
-
-	query, args := sb.BuildWithFlavor(sqlbuilder.ClickHouse)
-	return query, args, nil
-}
-
-// applySeriesConditions adds the WHERE conditions of a series table scan for
-// the given matchers and window. __name__ matchers translate to the
-// metric_name column (all four matcher types — the v1 client silently
-// returned nothing for regex metric names); every other matcher translates
-// to a JSONExtractString condition on the labels column. An equality matcher
-// against "" matches series without the label, mirroring PromQL, because
-// JSONExtractString returns "" for missing keys. Regexes are anchored:
-// PromQL matchers match the whole value, while ClickHouse match() searches
-// for a partial match — without anchoring, =~"api" would also select
-// "x-api-y".
-func applySeriesConditions(sb *sqlbuilder.SelectBuilder, start, end int64, matchers []*labels.Matcher) error {
-	for _, m := range matchers {
-		if m.Name != metricNameLabel {
-			continue
-		}
-		switch m.Type {
-		case labels.MatchEqual:
-			sb.Where(sb.EQ("metric_name", m.Value))
-		case labels.MatchNotEqual:
-			sb.Where(sb.NE("metric_name", m.Value))
-		case labels.MatchRegexp:
-			sb.Where(fmt.Sprintf("match(metric_name, %s)", sb.Var(anchorRegex(m.Value))))
-		case labels.MatchNotRegexp:
-			sb.Where(fmt.Sprintf("NOT match(metric_name, %s)", sb.Var(anchorRegex(m.Value))))
-		default:
-			return errors.NewInvalidInputf(errors.CodeInvalidInput, "unsupported matcher type %q for __name__", m.Type)
-		}
-	}
-
-	sb.Where("temporality IN ['Cumulative', 'Unspecified']")
-	sb.Where(fmt.Sprintf("__normalized = %v", !constants.IsDotMetricsEnabled))
-	sb.Where(sb.GTE("unix_milli", start), sb.LT("unix_milli", end))
-
-	for _, m := range matchers {
-		if m.Name == metricNameLabel {
-			continue
-		}
-		switch m.Type {
-		case labels.MatchEqual:
-			sb.Where(fmt.Sprintf("JSONExtractString(labels, %s) = %s", sb.Var(m.Name), sb.Var(m.Value)))
-		case labels.MatchNotEqual:
-			sb.Where(fmt.Sprintf("JSONExtractString(labels, %s) != %s", sb.Var(m.Name), sb.Var(m.Value)))
-		case labels.MatchRegexp:
-			sb.Where(fmt.Sprintf("match(JSONExtractString(labels, %s), %s)", sb.Var(m.Name), sb.Var(anchorRegex(m.Value))))
-		case labels.MatchNotRegexp:
-			sb.Where(fmt.Sprintf("NOT match(JSONExtractString(labels, %s), %s)", sb.Var(m.Name), sb.Var(anchorRegex(m.Value))))
-		default:
-			return errors.NewInvalidInputf(errors.CodeInvalidInput, "unsupported matcher type %q", m.Type)
-		}
-	}
-
-	return nil
-}
-
-// anchorRegex turns a PromQL regex into its fully-anchored form (see
-// applySeriesConditions).
-func anchorRegex(v string) string {
-	return "^(?:" + v + ")$"
-}
-
-// inlineFingerprintFilter renders "IN (fp1, fp2, ...)" with literal uint64s.
-func inlineFingerprintFilter(fingerprints []uint64) string {
-	var b strings.Builder
-	b.Grow(len(fingerprints)*21 + 8)
-	b.WriteString("IN (")
-	for i, fp := range fingerprints {
-		if i > 0 {
-			b.WriteString(", ")
-		}
-		b.WriteString(strconv.FormatUint(fp, 10))
-	}
-	b.WriteString(")")
-	return b.String()
-}
-
-// lastSamplePerStep reduces an instant-selector fetch to the last sample of
-// each step bucket. Buckets are anchored at the selector's first evaluation
-// timestamp so that every bucket boundary coincides with an evaluation
-// timestamp: bucket 0 is (start, firstEval] (the initial lookback window)
-// and bucket i is (firstEval+(i-1)·step, firstEval+i·step]. Keeping only the
-// last sample per bucket is lossless: the engine resolves each evaluation
-// timestamp t to the latest sample in (t-lookback, t], and a non-final
-// sample of a bucket can never be that latest sample for any t on the
-// evaluation grid. Real timestamps are preserved, so the engine's own
-// lookback and staleness handling remain exact.
-type lastSamplePerStep struct {
-	firstEvalMs int64
-	stepMs      int64
-}
-
-func (t *lastSamplePerStep) bucketExpr() string {
-	if t.stepMs <= 0 {
-		// Instant query: a single evaluation at firstEval; one bucket.
-		return ""
-	}
-	return fmt.Sprintf(
-		"if(unix_milli <= %d, 0, intDiv(unix_milli - %d - 1, %d) + 1)",
-		t.firstEvalMs, t.firstEvalMs, t.stepMs,
-	)
-}

pkg/prometheus/clickhouseprometheusv2/sql_test.go

@@ -1,148 +0,0 @@
-package clickhouseprometheusv2
-
-import (
-	"testing"
-	"time"
-
-	"github.com/prometheus/prometheus/model/labels"
-	"github.com/stretchr/testify/assert"
-	"github.com/stretchr/testify/require"
-)
-
-func mustMatcher(t *testing.T, mt labels.MatchType, name, value string) *labels.Matcher {
-	t.Helper()
-	m, err := labels.NewMatcher(mt, name, value)
-	require.NoError(t, err)
-	return m
-}
-
-func TestTimeSeriesTableFor(t *testing.T) {
-	base := time.Date(2026, 7, 10, 3, 27, 0, 0, time.UTC).UnixMilli()
-
-	tests := []struct {
-		name      string
-		span      time.Duration
-		wantTable string
-		roundTo   time.Duration
-	}{
-		{"under 6h uses hourly table", 2 * time.Hour, distributedTimeSeriesV4, time.Hour},
-		{"under 1d uses 6h table", 12 * time.Hour, distributedTimeSeriesV46hrs, 6 * time.Hour},
-		{"under 1w uses 1d table", 3 * 24 * time.Hour, distributedTimeSeriesV41day, 24 * time.Hour},
-		{"over 1w uses 1w table", 10 * 24 * time.Hour, distributedTimeSeriesV41week, 7 * 24 * time.Hour},
-	}
-	for _, tt := range tests {
-		t.Run(tt.name, func(t *testing.T) {
-			start, table := timeSeriesTableFor(base, base+tt.span.Milliseconds())
-			assert.Equal(t, tt.wantTable, table)
-			assert.Zero(t, start%tt.roundTo.Milliseconds())
-			assert.LessOrEqual(t, start, base)
-		})
-	}
-}
-
-func TestBuildSeriesQuery(t *testing.T) {
-	start := int64(1_700_000_000_000)
-	end := start + time.Hour.Milliseconds()
-	// The series table window rounds down to the table's bucket boundary.
-	adjustedStart := start - (start % time.Hour.Milliseconds())
-
-	t.Run("equality name and label matchers", func(t *testing.T) {
-		query, args, err := buildSeriesQuery(start, end, []*labels.Matcher{
-			mustMatcher(t, labels.MatchEqual, "__name__", "http_requests_total"),
-			mustMatcher(t, labels.MatchEqual, "job", "api"),
-		})
-		require.NoError(t, err)
-		assert.Equal(t,
-			"SELECT fingerprint, any(labels) FROM signoz_metrics.distributed_time_series_v4 WHERE metric_name = ? AND temporality IN ['Cumulative', 'Unspecified'] AND __normalized = false AND unix_milli >= ? AND unix_milli < ? AND JSONExtractString(labels, ?) = ? GROUP BY fingerprint",
-			query,
-		)
-		assert.Equal(t, []any{"http_requests_total", adjustedStart, end, "job", "api"}, args)
-	})
-
-	t.Run("regex matchers are anchored", func(t *testing.T) {
-		_, args, err := buildSeriesQuery(start, end, []*labels.Matcher{
-			mustMatcher(t, labels.MatchEqual, "__name__", "up"),
-			mustMatcher(t, labels.MatchRegexp, "instance", "prod.*"),
-			mustMatcher(t, labels.MatchNotRegexp, "env", "dev|test"),
-		})
-		require.NoError(t, err)
-		assert.Equal(t, []any{"up", adjustedStart, end, "instance", "^(?:prod.*)$", "env", "^(?:dev|test)$"}, args)
-	})
-
-	t.Run("regex name matcher uses metric_name column", func(t *testing.T) {
-		query, args, err := buildSeriesQuery(start, end, []*labels.Matcher{
-			mustMatcher(t, labels.MatchRegexp, "__name__", "node_cpu.*|node_memory.*"),
-		})
-		require.NoError(t, err)
-		assert.Contains(t, query, "match(metric_name, ?)")
-		assert.NotContains(t, query, "JSONExtractString")
-		assert.Equal(t, []any{"^(?:node_cpu.*|node_memory.*)$", adjustedStart, end}, args)
-	})
-
-	t.Run("no name matcher omits metric_name condition", func(t *testing.T) {
-		query, _, err := buildSeriesQuery(start, end, []*labels.Matcher{
-			mustMatcher(t, labels.MatchEqual, "job", "api"),
-		})
-		require.NoError(t, err)
-		assert.NotContains(t, query, "metric_name")
-	})
-}
-
-func TestBuildSamplesQuery(t *testing.T) {
-	start := int64(1_700_000_000_000)
-	end := start + time.Hour.Milliseconds()
-	adjustedStart := start - (start % time.Hour.Milliseconds())
-	matchers := []*labels.Matcher{
-		mustMatcher(t, labels.MatchEqual, "__name__", "up"),
-		mustMatcher(t, labels.MatchEqual, "job", "api"),
-	}
-
-	t.Run("raw with inline fingerprints", func(t *testing.T) {
-		query, args, err := buildSamplesQuery(start, end, []string{"up"}, []uint64{7, 42}, matchers, nil)
-		require.NoError(t, err)
-		assert.Equal(t,
-			"SELECT fingerprint, unix_milli, value, flags FROM signoz_metrics.distributed_samples_v4 WHERE metric_name = ? AND temporality IN ['Cumulative', 'Unspecified'] AND fingerprint IN (7, 42) AND unix_milli >= ? AND unix_milli <= ? ORDER BY fingerprint, unix_milli",
-			query,
-		)
-		assert.Equal(t, []any{"up", start, end}, args)
-	})
-
-	t.Run("last-sample-per-step groups by step bucket anchored at first eval", func(t *testing.T) {
-		lastPerStep := &lastSamplePerStep{firstEvalMs: start + 299_999, stepMs: 60_000}
-		query, _, err := buildSamplesQuery(start, end, []string{"up"}, []uint64{7}, matchers, lastPerStep)
-		require.NoError(t, err)
-		assert.Contains(t, query, "argMax(value, unix_milli) AS val")
-		assert.Contains(t, query, "argMax(flags, unix_milli) AS fl")
-		assert.Contains(t, query, "GROUP BY fingerprint, if(unix_milli <= 1700000299999, 0, intDiv(unix_milli - 1700000299999 - 1, 60000) + 1)")
-		assert.Contains(t, query, "ORDER BY fingerprint, ts")
-		// Aliases must not shadow the source columns referenced in WHERE.
-		assert.NotContains(t, query, "AS unix_milli")
-		assert.NotContains(t, query, "AS value")
-		assert.NotContains(t, query, "AS flags")
-	})
-
-	t.Run("instant query keeps one bucket", func(t *testing.T) {
-		lastPerStep := &lastSamplePerStep{firstEvalMs: end, stepMs: 0}
-		query, _, err := buildSamplesQuery(start, end, []string{"up"}, []uint64{7}, matchers, lastPerStep)
-		require.NoError(t, err)
-		assert.Contains(t, query, "GROUP BY fingerprint ORDER BY fingerprint, ts")
-		assert.NotContains(t, query, "intDiv")
-	})
-
-	t.Run("over-limit set becomes a shard-local semi-join", func(t *testing.T) {
-		query, args, err := buildSamplesQuery(start, end, []string{"up"}, nil, matchers, nil)
-		require.NoError(t, err)
-		assert.Contains(t, query, "fingerprint IN (SELECT fingerprint FROM signoz_metrics.time_series_v4 WHERE ")
-		assert.NotContains(t, query, "GLOBAL IN")
-		// Args follow placeholder order: samples metric name, the semi-join's
-		// series predicates, then the samples window bounds.
-		assert.Equal(t, []any{"up", "up", adjustedStart, end, "job", "api", start, end}, args)
-	})
-
-	t.Run("multiple metric names from regex selector", func(t *testing.T) {
-		query, args, err := buildSamplesQuery(start, end, []string{"node_cpu", "node_memory"}, []uint64{7}, matchers, nil)
-		require.NoError(t, err)
-		assert.Contains(t, query, "metric_name IN (?, ?)")
-		assert.Equal(t, []any{"node_cpu", "node_memory", start, end}, args)
-	})
-}

pkg/prometheus/clickhouseprometheusv2/tables.go

@@ -1,65 +0,0 @@
-package clickhouseprometheusv2
-
-import "time"
-
-const (
-	// metricNameLabel is the reserved PromQL label holding the metric name.
-	metricNameLabel string = "__name__"
-
-	databaseName                 string = "signoz_metrics"
-	distributedTimeSeriesV4      string = "distributed_time_series_v4"
-	distributedTimeSeriesV46hrs  string = "distributed_time_series_v4_6hrs"
-	distributedTimeSeriesV41day  string = "distributed_time_series_v4_1day"
-	distributedTimeSeriesV41week string = "distributed_time_series_v4_1week"
-	distributedSamplesV4         string = "distributed_samples_v4"
-
-	localTimeSeriesV4      string = "time_series_v4"
-	localTimeSeriesV46hrs  string = "time_series_v4_6hrs"
-	localTimeSeriesV41day  string = "time_series_v4_1day"
-	localTimeSeriesV41week string = "time_series_v4_1week"
-)
-
-// localTimeSeriesTable maps a distributed time series table to its shard-local
-// table. Samples and time series shard on the same key
-// (cityHash64(env, temporality, metric_name, fingerprint)), so a query whose
-// top-level FROM is the distributed samples table can join or semi-join the
-// local time series table inside each shard: the shard rewrite runs the
-// subquery against the shard's own series rows, which are exactly the series
-// of the shard's samples. No broadcast, no initiator-side join.
-func localTimeSeriesTable(distributed string) string {
-	switch distributed {
-	case distributedTimeSeriesV46hrs:
-		return localTimeSeriesV46hrs
-	case distributedTimeSeriesV41day:
-		return localTimeSeriesV41day
-	case distributedTimeSeriesV41week:
-		return localTimeSeriesV41week
-	default:
-		return localTimeSeriesV4
-	}
-}
-
-var (
-	oneHourInMilliseconds  = time.Hour.Milliseconds()
-	sixHoursInMilliseconds = time.Hour.Milliseconds() * 6
-	oneDayInMilliseconds   = time.Hour.Milliseconds() * 24
-	oneWeekInMilliseconds  = time.Hour.Milliseconds() * 24 * 7
-)
-
-// timeSeriesTableFor returns the adjusted start and the time series table for
-// the window. Time series tables hold one row per (fingerprint, bucket), with
-// bucket granularities of 1h, 6h, 1d and 1w; the start is rounded down to the
-// bucket boundary so a window beginning mid-bucket still matches the bucket's
-// row.
-func timeSeriesTableFor(start, end int64) (int64, string) {
-	switch {
-	case end-start < sixHoursInMilliseconds:
-		return start - (start % oneHourInMilliseconds), distributedTimeSeriesV4
-	case end-start < oneDayInMilliseconds:
-		return start - (start % sixHoursInMilliseconds), distributedTimeSeriesV46hrs
-	case end-start < oneWeekInMilliseconds:
-		return start - (start % oneDayInMilliseconds), distributedTimeSeriesV41day
-	default:
-		return start - (start % oneWeekInMilliseconds), distributedTimeSeriesV41week
-	}
-}

pkg/prometheus/clickhouseprometheusv2/transpiler.go

@@ -1,485 +0,0 @@
-package clickhouseprometheusv2
-
-import (
-	"fmt"
-	"strings"
-
-	"github.com/prometheus/prometheus/model/labels"
-	"github.com/prometheus/prometheus/promql/parser"
-)
-
-// The compiler turns PromQL subtrees into single ClickHouse queries built on
-// the timeSeries*ToGrid aggregate functions (CH >= 25.6), whose semantics
-// were verified against this repo's vendored engine: exact extrapolatedRate
-// behavior including counter resets, the counter zero-point clamp, the
-// 1.1x-average extrapolation threshold, left-open windows, the >= 2 samples
-// rule, stale-marker shadowing, and millisecond grid starts. Sample rows
-// never leave ClickHouse: one row per output series comes back, holding the
-// whole grid as an array.
-//
-// Scope (the allowlist): an optional sum/min/max/avg/count by/without
-// aggregation over a core unit — a rate/increase/delta/irate/idelta range
-// selection, an instant vector selection, or an avg/min/max/sum/count/last
-// _over_time window — plus number-literal arithmetic/comparisons and unary
-// minus on top. Units inside fixed-resolution subqueries evaluate on the
-// subquery's own grid. Everything else either falls back to the engine over
-// this package's querier, or — when a transpilable subtree sits under a
-// non-transpilable node — runs hybrid: the subtree's grids are computed in
-// ClickHouse and substituted into the engine as synthetic series (see
-// compiler_exec.go). See doc.go for the fallback list and the reasons behind
-// each entry.
-
-// rangeFn is a transpilable range-vector function.
-type rangeFn string
-
-const (
-	fnRate     rangeFn = "rate"
-	fnIncrease rangeFn = "increase"
-	fnDelta    rangeFn = "delta"
-	fnIRate    rangeFn = "irate"
-	fnIDelta   rangeFn = "idelta"
-)
-
-var gridFunction = map[rangeFn]string{
-	fnRate:     "timeSeriesRateToGrid",
-	fnIncrease: "timeSeriesRateToGrid", // increase == rate * range seconds, exactly (same factor algebra)
-	fnDelta:    "timeSeriesDeltaToGrid",
-	fnIRate:    "timeSeriesInstantRateToGrid",
-	fnIDelta:   "timeSeriesInstantDeltaToGrid",
-}
-
-// scalarOp is one number-literal arithmetic or comparison applied to a
-// compiled vector, evaluated in Go during assembly with the same float64
-// operations the engine uses.
-type scalarOp struct {
-	op           parser.ItemType
-	scalar       float64
-	scalarOnLeft bool
-	returnBool   bool
-}
-
-// isComparison reports whether the op is a filtering/bool comparison, which
-// preserves the metric name (arithmetic drops it).
-func (o scalarOp) isComparison() bool {
-	return o.op.IsComparisonOperator()
-}
-
-// unitKind is the selector shape at the bottom of a core unit.
-type unitKind int
-
-const (
-	// unitRange: rate/increase/delta/irate/idelta over a matrix selector.
-	unitRange unitKind = iota
-	// unitInstant: a plain vector selector resolved per grid point with
-	// lookback and stale-marker shadowing.
-	unitInstant
-	// unitOverTime: avg/min/max/sum/count/last_over_time over a matrix
-	// selector (aggregation over the window's samples, stale rows excluded).
-	unitOverTime
-)
-
-// coreUnit is one transpilable subtree: selector [-> range function] ->
-// optional aggregation -> scalar op pipeline.
-type coreUnit struct {
-	kind     unitKind
-	matchers []*labels.Matcher
-	offsetMs int64
-	fn       rangeFn // unitRange
-	overFn   string  // unitOverTime: avg|min|max|sum|count|last
-	rangeMs  int64   // unitRange/unitOverTime window
-
-	hasAgg   bool
-	aggOp    parser.ItemType // SUM MIN MAX AVG COUNT
-	by       bool
-	grouping []string
-
-	ops []scalarOp
-}
-
-// keepsName reports whether the unit's output series keep their real
-// __name__: bare/comparison-filtered instant selectors and last_over_time do
-// (it returns the raw sample, name included); range functions, the other
-// *_over_time functions, aggregations, arithmetic and bool comparisons all
-// drop it — a bool comparison returns 0/1, not the sample, so the engine
-// drops the name there too. Units that keep the name cannot be substituted
-// as synthetic series in hybrid plans — the synthetic name would replace
-// the real one — but transpile fine as full plans, where assembly emits the
-// real names.
-func (u *coreUnit) keepsName() bool {
-	nameKeepingSelector := u.kind == unitInstant || (u.kind == unitOverTime && u.overFn == "last")
-	if !nameKeepingSelector || u.hasAgg {
-		return false
-	}
-	for _, op := range u.ops {
-		if !op.isComparison() || op.returnBool {
-			return false
-		}
-	}
-	return true
-}
-
-// gridContext is the evaluation grid a unit computes on. The query grid for
-// top-level units; for units inside subqueries, the subquery's own grid:
-// epoch-aligned multiples of its resolution covering the subquery window,
-// exactly as the engine derives it (engine.go, *parser.SubqueryExpr case).
-type gridContext struct {
-	startMs int64
-	endMs   int64
-	stepMs  int64
-}
-
-// subqueryGrid derives the inner grid for a subquery evaluated on outer:
-// interval S, end = outer end − offset, start = first multiple of S strictly
-// greater than outer start − offset − range.
-func subqueryGrid(outer gridContext, rangeMs, stepMs, offsetMs int64) gridContext {
-	lower := outer.startMs - offsetMs - rangeMs
-	start := stepMs * (lower / stepMs)
-	if start <= lower {
-		start += stepMs
-	}
-	return gridContext{startMs: start, endMs: outer.endMs - offsetMs, stepMs: stepMs}
-}
-
-// transpiledUnit is a coreUnit scheduled for execution, named for hybrid
-// substitution, carrying the grid it evaluates on.
-type transpiledUnit struct {
-	core coreUnit
-	name string // __signoz_transpiled_<n>__
-	grid gridContext
-}
-
-// transpilePlan is the outcome of classifying a query.
-type transpilePlan struct {
-	units []*transpiledUnit
-	grid  gridContext // the query's top-level grid
-	// full is set when the entire query is units[0]; otherwise rewritten
-	// holds the query with each unit replaced by a synthetic selector, to be
-	// evaluated by the engine over a hybrid storage.
-	full      bool
-	rewritten string
-}
-
-const syntheticNamePrefix = "__signoz_transpiled_"
-
-func syntheticName(i int) string {
-	return fmt.Sprintf("%s%d__", syntheticNamePrefix, i)
-}
-
-// classifyCore matches a subtree against the transpilable core shape.
-// stepMs gates second-granularity: the grid functions take whole-second step
-// and window parameters (grid *starts* are millisecond-precise).
-func classifyCore(node parser.Expr, stepMs int64) (*coreUnit, bool) {
-	unit := &coreUnit{}
-
-	expr := node
-	// Peel scalar ops and parens off the top, outermost first; ops apply in
-	// evaluation order, so prepend while peeling.
-	for {
-		switch n := expr.(type) {
-		case *parser.ParenExpr:
-			expr = n.Expr
-			continue
-		case *parser.UnaryExpr:
-			if n.Op != parser.SUB {
-				expr = n.Expr // unary '+' is a no-op
-				continue
-			}
-			// -x == -1 * x for every float64 (incl. NaN and signed zero).
-			unit.ops = append([]scalarOp{{op: parser.MUL, scalar: -1}}, unit.ops...)
-			expr = n.Expr
-			continue
-		case *parser.StepInvariantExpr:
-			// @-pinned expressions evaluate on a different grid.
-			return nil, false
-		case *parser.BinaryExpr:
-			lit, litOnLeft, ok := numberLiteralSide(n)
-			if !ok {
-				return nil, false
-			}
-			if !n.Op.IsOperator() && !n.Op.IsComparisonOperator() {
-				return nil, false
-			}
-			if n.Op == parser.ATAN2 {
-				// atan2 is arithmetic in PromQL but rarely used; keep the
-				// allowlist tight.
-				return nil, false
-			}
-			returnBool := n.ReturnBool
-			unit.ops = append([]scalarOp{{op: n.Op, scalar: lit, scalarOnLeft: litOnLeft, returnBool: returnBool}}, unit.ops...)
-			if litOnLeft {
-				expr = n.RHS
-			} else {
-				expr = n.LHS
-			}
-			continue
-		}
-		break
-	}
-
-	// Optional aggregation.
-	if agg, ok := expr.(*parser.AggregateExpr); ok {
-		switch agg.Op {
-		case parser.SUM, parser.MIN, parser.MAX, parser.AVG, parser.COUNT:
-		default:
-			return nil, false
-		}
-		for _, g := range agg.Grouping {
-			if g == metricNameLabel {
-				// by(__name__)/without(__name__) over synthetic or compiled
-				// output needs name bookkeeping the compiler doesn't do.
-				return nil, false
-			}
-		}
-		unit.hasAgg = true
-		unit.aggOp = agg.Op
-		unit.by = !agg.Without
-		unit.grouping = agg.Grouping
-		expr = agg.Expr
-		for {
-			if p, ok := expr.(*parser.ParenExpr); ok {
-				expr = p.Expr
-				continue
-			}
-			break
-		}
-	}
-
-	// The grid functions take whole-second steps; stepMs == 0 is an instant
-	// query (single-point grid).
-	if stepMs < 0 || stepMs%1000 != 0 {
-		return nil, false
-	}
-
-	// Bare instant selector: resolved per grid point with lookback and
-	// stale-marker shadowing (see compiler_sql.go).
-	if vs, ok := expr.(*parser.VectorSelector); ok {
-		if vs.Timestamp != nil || vs.StartOrEnd != 0 || vs.Anchored || vs.Smoothed {
-			return nil, false
-		}
-		offsetMs := vs.OriginalOffset.Milliseconds()
-		if offsetMs < 0 {
-			return nil, false
-		}
-		unit.kind = unitInstant
-		unit.offsetMs = offsetMs
-		unit.matchers = vs.LabelMatchers
-		return unit, true
-	}
-
-	// Range or *_over_time function over a plain matrix selector.
-	call, ok := expr.(*parser.Call)
-	if !ok {
-		return nil, false
-	}
-	var fn rangeFn
-	var overFn string
-	switch call.Func.Name {
-	case "rate":
-		fn = fnRate
-	case "increase":
-		fn = fnIncrease
-	case "delta":
-		fn = fnDelta
-	case "irate":
-		fn = fnIRate
-	case "idelta":
-		fn = fnIDelta
-	case "avg_over_time", "min_over_time", "max_over_time", "sum_over_time", "count_over_time", "last_over_time":
-		overFn = strings.TrimSuffix(call.Func.Name, "_over_time")
-	default:
-		return nil, false
-	}
-	if len(call.Args) != 1 {
-		return nil, false
-	}
-	ms, ok := call.Args[0].(*parser.MatrixSelector)
-	if !ok {
-		return nil, false
-	}
-	vs, ok := ms.VectorSelector.(*parser.VectorSelector)
-	if !ok {
-		return nil, false
-	}
-	if vs.Timestamp != nil || vs.StartOrEnd != 0 || vs.Anchored || vs.Smoothed {
-		return nil, false
-	}
-
-	rangeMs := ms.Range.Milliseconds()
-	offsetMs := vs.OriginalOffset.Milliseconds()
-	if rangeMs <= 0 || rangeMs%1000 != 0 || offsetMs < 0 {
-		return nil, false
-	}
-
-	if overFn != "" {
-		unit.kind = unitOverTime
-		unit.overFn = overFn
-	} else {
-		unit.kind = unitRange
-		unit.fn = fn
-	}
-	unit.rangeMs = rangeMs
-	unit.offsetMs = offsetMs
-	unit.matchers = vs.LabelMatchers
-	return unit, true
-}
-
-// numberLiteralSide returns the number literal on one side of a binary
-// expression (peeling parens and unary minus), and which side it is on.
-func numberLiteralSide(b *parser.BinaryExpr) (float64, bool, bool) {
-	if v, ok := literalValue(b.LHS); ok {
-		return v, true, true
-	}
-	if v, ok := literalValue(b.RHS); ok {
-		return v, false, true
-	}
-	return 0, false, false
-}
-
-func literalValue(e parser.Expr) (float64, bool) {
-	neg := false
-	for {
-		switch n := e.(type) {
-		case *parser.ParenExpr:
-			e = n.Expr
-			continue
-		case *parser.StepInvariantExpr:
-			e = n.Expr
-			continue
-		case *parser.UnaryExpr:
-			if n.Op == parser.SUB {
-				neg = !neg
-			}
-			e = n.Expr
-			continue
-		case *parser.NumberLiteral:
-			if neg {
-				return -n.Val, true
-			}
-			return n.Val, true
-		default:
-			return 0, false
-		}
-	}
-}
-
-// classify builds the compile plan for a query: full when the root is a core
-// unit, hybrid when core units sit strictly below the root (including inside
-// fixed-resolution subqueries, computed on the subquery grid), none
-// otherwise.
-func classify(root parser.Expr, grid gridContext) (*transpilePlan, bool) {
-	if unit, ok := classifyCore(root, grid.stepMs); ok {
-		return &transpilePlan{
-			units: []*transpiledUnit{{core: *unit, name: syntheticName(0), grid: grid}},
-			grid:  grid,
-			full:  true,
-		}, true
-	}
-
-	plan := &transpilePlan{grid: grid}
-	rewritten := rewrite(root, grid, plan, false)
-	if len(plan.units) == 0 {
-		return nil, false
-	}
-	plan.rewritten = rewritten.String()
-	return plan, true
-}
-
-// rewrite walks top-down replacing maximal transpilable subtrees with synthetic
-// vector selectors. nameSensitive marks scopes where an ancestor's semantics
-// depend on __name__ (grouping or vector matching on it): synthetic series
-// carry a synthetic __name__, so substitution there would change results.
-// Fixed-resolution subqueries recurse with the subquery's own grid; scopes
-// whose evaluation grid is unknowable (@-pinned, default-resolution
-// subqueries) are not entered.
-func rewrite(node parser.Expr, grid gridContext, plan *transpilePlan, nameSensitive bool) parser.Expr {
-	if node == nil {
-		return nil
-	}
-
-	if !nameSensitive {
-		// Units whose output keeps the real __name__ (bare instant selectors)
-		// cannot be substituted: the synthetic name would replace it in the
-		// engine's output. They still compile as full plans.
-		if unit, ok := classifyCore(node, grid.stepMs); ok && !unit.keepsName() {
-			cu := &transpiledUnit{core: *unit, name: syntheticName(len(plan.units)), grid: grid}
-			plan.units = append(plan.units, cu)
-			return &parser.VectorSelector{
-				Name: cu.name,
-				LabelMatchers: []*labels.Matcher{
-					labels.MustNewMatcher(labels.MatchEqual, metricNameLabel, cu.name),
-				},
-				PosRange: node.PositionRange(),
-			}
-		}
-	}
-
-	switch n := node.(type) {
-	case *parser.ParenExpr:
-		n.Expr = rewrite(n.Expr, grid, plan, nameSensitive)
-	case *parser.UnaryExpr:
-		n.Expr = rewrite(n.Expr, grid, plan, nameSensitive)
-	case *parser.AggregateExpr:
-		sensitive := nameSensitive || groupingUsesName(n.Grouping)
-		n.Expr = rewrite(n.Expr, grid, plan, sensitive)
-		// n.Param is a scalar/string; nothing transpilable inside for our core.
-	case *parser.Call:
-		for i, arg := range n.Args {
-			n.Args[i] = rewrite(arg, grid, plan, nameSensitive)
-		}
-	case *parser.BinaryExpr:
-		sensitive := nameSensitive || vectorMatchingUsesName(n.VectorMatching)
-		n.LHS = rewrite(n.LHS, grid, plan, sensitive)
-		n.RHS = rewrite(n.RHS, grid, plan, sensitive)
-	case *parser.SubqueryExpr:
-		// The alert-smoothing idiom fn_over_time((expr)[R:S]) dominates real
-		// rule fleets; inner units evaluate on the subquery grid, and the
-		// engine does the smoothing over the synthetic series. Requires an
-		// explicit whole-second resolution (S == 0 needs the engine's
-		// default-interval function) and no @ pinning.
-		stepMs := n.Step.Milliseconds()
-		rangeMs := n.Range.Milliseconds()
-		offsetMs := n.OriginalOffset.Milliseconds()
-		if n.Timestamp == nil && n.StartOrEnd == 0 &&
-			stepMs > 0 && stepMs%1000 == 0 && rangeMs%1000 == 0 && offsetMs >= 0 {
-			inner := subqueryGrid(grid, rangeMs, stepMs, offsetMs)
-			n.Expr = rewrite(n.Expr, inner, plan, nameSensitive)
-		}
-	case *parser.StepInvariantExpr, *parser.MatrixSelector,
-		*parser.VectorSelector, *parser.NumberLiteral, *parser.StringLiteral:
-		// Leaves, or scopes substitution must not enter.
-	}
-	return node
-}
-
-func groupingUsesName(grouping []string) bool {
-	for _, g := range grouping {
-		if g == metricNameLabel {
-			return true
-		}
-	}
-	return false
-}
-
-func vectorMatchingUsesName(vm *parser.VectorMatching) bool {
-	if vm == nil {
-		return false
-	}
-	for _, l := range append(append([]string{}, vm.MatchingLabels...), vm.Include...) {
-		if l == metricNameLabel {
-			return true
-		}
-	}
-	// Default (all-labels) matching ignores __name__, and by()/ignoring()
-	// lists were checked above.
-	return false
-}
-
-// isSyntheticSelector reports whether matchers target a compiled unit.
-func isSyntheticSelector(matchers []*labels.Matcher) (string, bool) {
-	for _, m := range matchers {
-		if m.Name == metricNameLabel && m.Type == labels.MatchEqual && strings.HasPrefix(m.Value, syntheticNamePrefix) {
-			return m.Value, true
-		}
-	}
-	return "", false
-}

pkg/prometheus/clickhouseprometheusv2/transpiler_corpus_test.go

@@ -1,150 +0,0 @@
-package clickhouseprometheusv2
-
-import (
-	"bufio"
-	"encoding/json"
-	"fmt"
-	"os"
-	"regexp"
-	"sort"
-	"strings"
-	"testing"
-
-	"github.com/prometheus/prometheus/promql/parser"
-	"github.com/stretchr/testify/require"
-)
-
-// TestClassifyCorpus measures real-workload compiler coverage: it classifies
-// every query of a JSON-lines corpus (one JSON-encoded PromQL string per
-// line) with the live classifier and reports full / hybrid / fallback
-// shares. Skipped unless PROMQL_CORPUS points to one or more files
-// (comma-separated). Dashboard template variables are substituted with
-// placeholder values before parsing, mirroring the production render step.
-//
-//	PROMQL_CORPUS=corpus-a.jsonl,corpus-b.jsonl go test -run TestClassifyCorpus -v
-func TestClassifyCorpus(t *testing.T) {
-	corpus := os.Getenv("PROMQL_CORPUS")
-	if corpus == "" {
-		t.Skip("PROMQL_CORPUS not set")
-	}
-
-	varRe := regexp.MustCompile(`\{\{\s*\.?[\w.]+\s*\}\}|\[\[\s*[\w.]+\s*\]\]|\$[\w.]+`)
-	promParser := parser.NewParser(parser.Options{})
-
-	for _, path := range strings.Split(corpus, ",") {
-		f, err := os.Open(path)
-		require.NoError(t, err)
-
-		var full, hybrid, fallbackInstant, fallbackOther, parseErrs int
-		fallbackReasons := map[string]int{}
-
-		scanner := bufio.NewScanner(f)
-		scanner.Buffer(make([]byte, 1024*1024), 1024*1024)
-		for scanner.Scan() {
-			var query string
-			require.NoError(t, json.Unmarshal(scanner.Bytes(), &query))
-			query = varRe.ReplaceAllString(query, "placeholder")
-
-			expr, err := promParser.ParseExpr(query)
-			if err != nil {
-				parseErrs++
-				continue
-			}
-
-			plan, ok := classify(expr, gridContext{startMs: 1_700_000_000_000, endMs: 1_700_007_200_000, stepMs: 60_000})
-			switch {
-			case ok && plan.full:
-				full++
-			case ok:
-				hybrid++
-			default:
-				reason := fallbackShape(expr)
-				fallbackReasons[reason]++
-				if reason == "instant-selector shape (last-sample-per-step engine path)" {
-					fallbackInstant++
-				} else {
-					fallbackOther++
-				}
-			}
-		}
-		require.NoError(t, scanner.Err())
-		_ = f.Close()
-
-		total := full + hybrid + fallbackInstant + fallbackOther
-		if total == 0 {
-			t.Logf("%s: no parseable queries (%d parse errors)", path, parseErrs)
-			continue
-		}
-		t.Logf("%s: %d queries — full=%d (%.0f%%) hybrid=%d (%.0f%%) fallback=%d (%.0f%%; instant-shape=%d) parse_errors=%d",
-			path, total,
-			full, 100*float64(full)/float64(total),
-			hybrid, 100*float64(hybrid)/float64(total),
-			fallbackInstant+fallbackOther, 100*float64(fallbackInstant+fallbackOther)/float64(total),
-			fallbackInstant, parseErrs)
-
-		reasons := make([]string, 0, len(fallbackReasons))
-		for r := range fallbackReasons {
-			reasons = append(reasons, r)
-		}
-		sort.Slice(reasons, func(i, j int) bool { return fallbackReasons[reasons[i]] > fallbackReasons[reasons[j]] })
-		for _, r := range reasons {
-			t.Logf("  fallback %4d  %s", fallbackReasons[r], r)
-		}
-	}
-}
-
-// fallbackShape buckets a non-transpilable query by why it stays on the engine
-// path, to separate "already served well" (instant selectors on the last-sample-per-step
-// path) from genuine compiler gaps.
-func fallbackShape(expr parser.Expr) string {
-	var hasMatrix, hasSubquery, hasAt, overTime bool
-	rangeFns := map[string]bool{"rate": true, "increase": true, "delta": true, "irate": true, "idelta": true}
-	var unsupportedFns []string
-	parser.Inspect(expr, func(node parser.Node, _ []parser.Node) error {
-		switch n := node.(type) {
-		case *parser.MatrixSelector:
-			hasMatrix = true
-		case *parser.SubqueryExpr:
-			hasSubquery = true
-		case *parser.VectorSelector:
-			if n.Timestamp != nil || n.StartOrEnd != 0 {
-				hasAt = true
-			}
-		case *parser.Call:
-			if strings.HasSuffix(n.Func.Name, "_over_time") {
-				overTime = true
-			} else if !rangeFns[n.Func.Name] {
-				unsupportedFns = append(unsupportedFns, n.Func.Name)
-			}
-		}
-		return nil
-	})
-
-	switch {
-	case hasSubquery:
-		return "subquery"
-	case hasAt:
-		return "@ modifier"
-	case overTime:
-		return "*_over_time range function"
-	case !hasMatrix:
-		return "instant-selector shape (last-sample-per-step engine path)"
-	case len(unsupportedFns) > 0:
-		return fmt.Sprintf("range shape with unsupported function(s): %s", strings.Join(dedupe(unsupportedFns), ",")) //nolint:makezero
-	default:
-		return "other range shape"
-	}
-}
-
-func dedupe(in []string) []string {
-	seen := map[string]bool{}
-	var out []string
-	for _, s := range in {
-		if !seen[s] {
-			seen[s] = true
-			out = append(out, s)
-		}
-	}
-	sort.Strings(out)
-	return out
-}

pkg/prometheus/clickhouseprometheusv2/transpiler_exec.go

@@ -1,463 +0,0 @@
-package clickhouseprometheusv2
-
-import (
-	"context"
-	"encoding/json"
-	"math"
-	"sort"
-	"sync/atomic"
-	"time"
-
-	"github.com/SigNoz/signoz/pkg/errors"
-	"github.com/SigNoz/signoz/pkg/prometheus"
-	"github.com/prometheus/prometheus/model/labels"
-	promValue "github.com/prometheus/prometheus/model/value"
-	"github.com/prometheus/prometheus/promql"
-	"github.com/prometheus/prometheus/promql/parser"
-	"github.com/prometheus/prometheus/storage"
-	"golang.org/x/sync/errgroup"
-)
-
-// executor evaluates transpilable PromQL directly in ClickHouse, falling
-// back (ok=false) whenever the query shape or the step doesn't qualify. The
-// timeSeries*ToGrid functions it builds on are assumed available: the
-// supported ClickHouse floor is >= 25.6.
-type executor struct {
-	client *client
-	engine *prometheus.Engine
-	parser prometheus.Parser
-}
-
-// TryExecuteRange transpiles and runs the query in ClickHouse when its shape
-// is in the allowlist. ok=false means "not transpilable" and carries no
-// error; the caller runs the engine path.
-func (e *executor) TryExecuteRange(ctx context.Context, qs string, start, end time.Time, step time.Duration) (promql.Matrix, bool, error) {
-	expr, err := e.parser.ParseExpr(qs)
-	if err != nil {
-		// Let the engine path produce the (enhanced) parse error.
-		return nil, false, nil
-	}
-
-	plan, ok := classify(expr, queryGrid(start, end, step))
-	if !ok {
-		return nil, false, nil
-	}
-
-	// timeSeriesLastToGrid widens its window to max(window, step) — probed: a
-	// sample aged (window, step] still fills the slot — while the rate/delta
-	// family enforces the window strictly. The Last-style kinds therefore
-	// transpile only when their window covers the step; otherwise the engine
-	// path serves them exactly.
-	for _, unit := range plan.units {
-		lastStyle := unit.core.kind == unitInstant || (unit.core.kind == unitOverTime && unit.core.overFn == "last")
-		if !lastStyle {
-			continue
-		}
-		windowMs := unit.core.rangeMs
-		if unit.core.kind == unitInstant {
-			windowMs = e.client.lookbackMs
-		}
-		if windowMs < unit.grid.stepMs {
-			return nil, false, nil
-		}
-	}
-
-	// Evaluate every unit concurrently on its own grid (the query grid, or a
-	// subquery grid); each is one series lookup plus one grid query. The
-	// units share one grid-cell budget: transpiled results never pass
-	// through the engine's sample limiter, so without this a large
-	// series-count x grid-width query would buffer unbounded arrays — the
-	// OOM this provider exists to prevent.
-	results := make([][]transpiledSeries, len(plan.units))
-	var gridCells atomic.Int64
-	eg, egCtx := errgroup.WithContext(ctx)
-	for i, unit := range plan.units {
-		eg.Go(func() error {
-			res, err := e.executeUnit(egCtx, &unit.core, unit.grid, &gridCells)
-			if err != nil {
-				return err
-			}
-			results[i] = res
-			return nil
-		})
-	}
-	if err := eg.Wait(); err != nil {
-		return nil, true, err
-	}
-
-	if plan.full {
-		g := plan.units[0].grid
-		return toMatrix(results[0], g.startMs, g.stepMs), true, nil
-	}
-
-	matrix, err := e.executeHybrid(ctx, plan, results)
-	if err != nil {
-		return nil, true, err
-	}
-	return matrix, true, nil
-}
-
-// queryGrid derives the top-level evaluation grid; step 0 is an instant
-// query: a single evaluation at end, whatever start was.
-func queryGrid(start, end time.Time, step time.Duration) gridContext {
-	startMs, endMs, stepMs := start.UnixMilli(), end.UnixMilli(), step.Milliseconds()
-	if stepMs == 0 {
-		startMs = endMs
-	}
-	return gridContext{startMs: startMs, endMs: endMs, stepMs: stepMs}
-}
-
-// transpiledSeries is one output series of a unit: projected labels and one
-// value pointer per grid point (nil = absent).
-type transpiledSeries struct {
-	lset   labels.Labels
-	values []*float64
-}
-
-// executeUnit runs one core unit on its grid: series lookup (budgets,
-// fingerprints, metric names), then the single grid query, then the
-// scalar-op pipeline.
-func (e *executor) executeUnit(ctx context.Context, unit *coreUnit, grid gridContext, gridCells *atomic.Int64) ([]transpiledSeries, error) {
-	startMs, endMs, stepMs := grid.startMs, grid.endMs, grid.stepMs
-	windowMs := unit.rangeMs
-	if unit.kind == unitInstant {
-		windowMs = e.client.lookbackMs
-	}
-	dataStart := startMs - unit.offsetMs - windowMs
-	dataEnd := endMs - unit.offsetMs
-
-	seriesQuery, seriesArgs, err := buildSeriesQuery(dataStart, dataEnd, unit.matchers)
-	if err != nil {
-		return nil, err
-	}
-	lookup, err := e.client.selectSeries(ctx, seriesQuery, seriesArgs)
-	if err != nil {
-		return nil, err
-	}
-	if len(lookup.fingerprints) == 0 {
-		return nil, nil
-	}
-
-	// The result buffers one grid array per series; series count times grid
-	// width is the transpiled equivalent of fetched samples, counted before
-	// the arrays exist rather than after the memory is spent.
-	gridLen := int64(1)
-	if stepMs > 0 {
-		gridLen = (endMs-startMs)/stepMs + 1
-	}
-	if maxSamples := e.client.cfg.MaxFetchedSamples; maxSamples > 0 && gridCells.Add(int64(len(lookup.fingerprints))*gridLen) > maxSamples {
-		return nil, errors.NewInvalidInputf(
-			errors.CodeInvalidInput,
-			"promql query would buffer more than %d output points; narrow the selector or time range, or raise prometheus::clickhousev2::max_fetched_samples",
-			maxSamples,
-		)
-	}
-
-	query, args, err := buildUnitSQL(unit, lookup.metricNames, transpiledFingerprintFilter(lookup), dataStart, dataEnd, startMs, endMs, stepMs, e.client.lookbackMs)
-	if err != nil {
-		return nil, err
-	}
-
-	rows, err := e.client.telemetryStore.ClickhouseDB().Query(e.client.withContext(ctx, "transpiledUnit"), query, args...)
-	if err != nil {
-		return nil, err
-	}
-	defer rows.Close()
-
-	// Name-dropping units keep __name__ in the SQL group key so distinct
-	// metrics never merge server-side; the name comes off here, and a
-	// post-strip collision is the engine's duplicate-labelset error — v1
-	// would have errored, so silently inventing a merged series would be a
-	// divergence.
-	stripName := !unit.hasAgg && !unit.keepsName()
-	seen := make(map[uint64]string)
-
-	var out []transpiledSeries
-	var gkey string
-	var gridValues []*float64
-	for rows.Next() {
-		if err := rows.Scan(&gkey, &gridValues); err != nil {
-			return nil, err
-		}
-		lset, err := labelsFromGroupKey(gkey)
-		if err != nil {
-			return nil, err
-		}
-		if stripName {
-			name := lset.Get(metricNameLabel)
-			lset = labels.NewBuilder(lset).Del(metricNameLabel).Labels()
-			if prev, ok := seen[lset.Hash()]; ok && prev != name {
-				return nil, errors.NewInvalidInputf(errors.CodeInvalidInput, "vector cannot contain metrics with the same labelset")
-			}
-			seen[lset.Hash()] = name
-		}
-		values := make([]*float64, len(gridValues))
-		copy(values, gridValues)
-		applyScalarOps(unit.ops, values)
-		out = append(out, transpiledSeries{lset: lset, values: values})
-	}
-	if err := rows.Err(); err != nil {
-		return nil, err
-	}
-
-	sort.Slice(out, func(i, j int) bool { return labels.Compare(out[i].lset, out[j].lset) < 0 })
-	return out, nil
-}
-
-// transpiledFingerprintFilter returns the matched fingerprints as a sorted
-// slice when they fit the inline limit — literals engage the samples primary
-// key, and sorting keeps the statement deterministic for logging and tests.
-// Over the limit it returns nil: the unit query's INNER JOIN against the
-// local series subquery restricts to exactly the matched fingerprints
-// already, and a semi-join on the same predicates would only rescan the
-// series table.
-func transpiledFingerprintFilter(lookup *seriesLookup) []uint64 {
-	if len(lookup.fingerprints) > inlineFingerprintsLimit {
-		return nil
-	}
-	fingerprints := make([]uint64, 0, len(lookup.fingerprints))
-	for fp := range lookup.fingerprints {
-		fingerprints = append(fingerprints, fp)
-	}
-	sort.Slice(fingerprints, func(i, j int) bool { return fingerprints[i] < fingerprints[j] })
-	return fingerprints
-}
-
-// labelsFromGroupKey parses the toJSONString'd sorted [key, value] pairs.
-func labelsFromGroupKey(gkey string) (labels.Labels, error) {
-	var pairs [][]string
-	if err := json.Unmarshal([]byte(gkey), &pairs); err != nil {
-		return labels.EmptyLabels(), errors.WrapInternalf(err, errors.CodeInternal, "malformed compiled group key %q", gkey)
-	}
-	builder := labels.NewScratchBuilder(len(pairs))
-	for _, p := range pairs {
-		if len(p) != 2 {
-			return labels.EmptyLabels(), errors.NewInternalf(errors.CodeInternal, "malformed compiled group key pair %q", gkey)
-		}
-		builder.Add(p[0], p[1])
-	}
-	builder.Sort()
-	return builder.Labels(), nil
-}
-
-// applyScalarOps applies the number-literal op pipeline in place, with the
-// same float64 arithmetic and comparison-filter semantics as the engine.
-func applyScalarOps(ops []scalarOp, values []*float64) {
-	for _, op := range ops {
-		for i, v := range values {
-			if v == nil {
-				continue
-			}
-			lhs, rhs := *v, op.scalar
-			if op.scalarOnLeft {
-				lhs, rhs = op.scalar, *v
-			}
-			switch op.op {
-			case parser.ADD:
-				res := lhs + rhs
-				values[i] = &res
-			case parser.SUB:
-				res := lhs - rhs
-				values[i] = &res
-			case parser.MUL:
-				res := lhs * rhs
-				values[i] = &res
-			case parser.DIV:
-				res := lhs / rhs
-				values[i] = &res
-			case parser.MOD:
-				res := math.Mod(lhs, rhs)
-				values[i] = &res
-			case parser.POW:
-				res := math.Pow(lhs, rhs)
-				values[i] = &res
-			default:
-				keep := compare(op.op, lhs, rhs)
-				switch {
-				case op.returnBool:
-					res := 0.0
-					if keep {
-						res = 1.0
-					}
-					values[i] = &res
-				case keep:
-					// Filter comparisons keep the vector-side value.
-					vec := *v
-					values[i] = &vec
-				default:
-					values[i] = nil
-				}
-			}
-		}
-	}
-}
-
-func compare(op parser.ItemType, lhs, rhs float64) bool {
-	switch op {
-	case parser.EQLC:
-		return lhs == rhs
-	case parser.NEQ:
-		return lhs != rhs
-	case parser.GTR:
-		return lhs > rhs
-	case parser.LSS:
-		return lhs < rhs
-	case parser.GTE:
-		return lhs >= rhs
-	case parser.LTE:
-		return lhs <= rhs
-	}
-	return false
-}
-
-// toMatrix converts a unit result to a promql matrix on the query grid.
-func toMatrix(series []transpiledSeries, startMs, stepMs int64) promql.Matrix {
-	matrix := make(promql.Matrix, 0, len(series))
-	for _, s := range series {
-		var floats []promql.FPoint
-		for i, v := range s.values {
-			if v == nil {
-				continue
-			}
-			floats = append(floats, promql.FPoint{T: startMs + int64(i)*stepMs, F: *v})
-		}
-		if len(floats) == 0 {
-			continue
-		}
-		matrix = append(matrix, promql.Series{Metric: s.lset, Floats: floats})
-	}
-	return matrix
-}
-
-// executeHybrid substitutes each unit's grids into the engine as synthetic
-// series and evaluates the rewritten query over a storage that serves
-// synthetic selectors from memory and everything else from the live querier.
-// Absent grid points become stale markers so the engine's lookback cannot
-// resurrect the previous grid point. Each unit's synthetic samples sit on its
-// own grid (query grid, or subquery grid for units inside subqueries).
-func (e *executor) executeHybrid(ctx context.Context, plan *transpilePlan, results [][]transpiledSeries) (promql.Matrix, error) {
-	synthetic := make(map[string][]*series, len(plan.units))
-	staleMarker := math.Float64frombits(promValue.StaleNaN)
-
-	queryGrid := plan.grid
-
-	for i, unit := range plan.units {
-		g := unit.grid
-		gridLen := 1
-		if g.stepMs > 0 {
-			gridLen = int((g.endMs-g.startMs)/g.stepMs) + 1
-		}
-		list := make([]*series, 0, len(results[i]))
-		for _, cs := range results[i] {
-			builder := labels.NewBuilder(cs.lset)
-			builder.Set(metricNameLabel, unit.name)
-			s := &series{lset: builder.Labels()}
-			s.ts = make([]int64, 0, gridLen)
-			s.vs = make([]float64, 0, gridLen)
-			for idx := 0; idx < gridLen; idx++ {
-				t := g.startMs + int64(idx)*g.stepMs
-				var v float64
-				if idx < len(cs.values) && cs.values[idx] != nil {
-					v = *cs.values[idx]
-				} else {
-					v = staleMarker
-				}
-				s.ts = append(s.ts, t)
-				s.vs = append(s.vs, v)
-			}
-			list = append(list, s)
-		}
-		synthetic[unit.name] = list
-	}
-
-	hybrid := &hybridQueryable{client: e.client, synthetic: synthetic}
-
-	var qry promql.Query
-	var err error
-	if queryGrid.stepMs == 0 {
-		qry, err = e.engine.NewInstantQuery(ctx, hybrid, nil, plan.rewritten, time.UnixMilli(queryGrid.endMs))
-	} else {
-		qry, err = e.engine.NewRangeQuery(ctx, hybrid, nil, plan.rewritten, time.UnixMilli(queryGrid.startMs), time.UnixMilli(queryGrid.endMs), time.Duration(queryGrid.stepMs)*time.Millisecond)
-	}
-	if err != nil {
-		return nil, err
-	}
-	defer qry.Close()
-
-	res := qry.Exec(ctx)
-	if res.Err != nil {
-		return nil, res.Err
-	}
-
-	matrix, err := resultToMatrix(res)
-	if err != nil {
-		return nil, err
-	}
-
-	// Deep-copy before Close returns the result's slices to the engine pool,
-	// and drop the synthetic __name__ that filter comparisons preserve.
-	out := make(promql.Matrix, 0, len(matrix))
-	for _, s := range matrix {
-		lset := s.Metric
-		if name := lset.Get(metricNameLabel); len(name) >= len(syntheticNamePrefix) && name[:len(syntheticNamePrefix)] == syntheticNamePrefix {
-			builder := labels.NewBuilder(lset)
-			builder.Del(metricNameLabel)
-			lset = builder.Labels()
-		}
-		floats := make([]promql.FPoint, len(s.Floats))
-		copy(floats, s.Floats)
-		out = append(out, promql.Series{Metric: lset.Copy(), Floats: floats})
-	}
-	sort.Slice(out, func(i, j int) bool { return labels.Compare(out[i].Metric, out[j].Metric) < 0 })
-	return out, nil
-}
-
-func resultToMatrix(res *promql.Result) (promql.Matrix, error) {
-	switch v := res.Value.(type) {
-	case promql.Matrix:
-		return v, nil
-	case promql.Vector:
-		matrix := make(promql.Matrix, 0, len(v))
-		for _, s := range v {
-			matrix = append(matrix, promql.Series{Metric: s.Metric, Floats: []promql.FPoint{{T: s.T, F: s.F}}})
-		}
-		return matrix, nil
-	case promql.Scalar:
-		return promql.Matrix{{Metric: labels.EmptyLabels(), Floats: []promql.FPoint{{T: v.T, F: v.V}}}}, nil
-	default:
-		return nil, errors.NewInternalf(errors.CodeInternal, "unexpected hybrid result type %T", res.Value)
-	}
-}
-
-// hybridQueryable serves synthetic (compiled) selectors from memory and
-// everything else from the live storage.
-type hybridQueryable struct {
-	client    *client
-	synthetic map[string][]*series
-}
-
-func (h *hybridQueryable) Querier(mint, maxt int64) (storage.Querier, error) {
-	return &hybridQuerier{
-		querier:   querier{mint: mint, maxt: maxt, client: h.client},
-		synthetic: h.synthetic,
-	}, nil
-}
-
-type hybridQuerier struct {
-	querier
-	synthetic map[string][]*series
-}
-
-func (h *hybridQuerier) Select(ctx context.Context, sortSeries bool, hints *storage.SelectHints, matchers ...*labels.Matcher) storage.SeriesSet {
-	if name, ok := isSyntheticSelector(matchers); ok {
-		list := h.synthetic[name]
-		if sortSeries {
-			sorted := make([]*series, len(list))
-			copy(sorted, list)
-			sort.Slice(sorted, func(i, j int) bool { return labels.Compare(sorted[i].lset, sorted[j].lset) < 0 })
-			list = sorted
-		}
-		return newSeriesSet(list)
-	}
-	return h.querier.Select(ctx, sortSeries, hints, matchers...)
-}

pkg/prometheus/clickhouseprometheusv2/transpiler_sql.go

@@ -1,303 +0,0 @@
-package clickhouseprometheusv2
-
-import (
-	"fmt"
-	"strings"
-
-	"github.com/huandu/go-sqlbuilder"
-)
-
-// experimental gate for the timeSeries*ToGrid aggregate functions; attached
-// as a SETTINGS clause so telemetrystore hooks cannot clobber it.
-const gridFunctionsSetting = "SETTINGS allow_experimental_ts_to_grid_aggregate_function = 1"
-
-var aggForEach = map[string]string{
-	"sum":   "sumForEach",
-	"min":   "minForEach",
-	"max":   "maxForEach",
-	"avg":   "avgForEach",
-	"count": "countForEach",
-}
-
-// buildUnitSQL renders the single ClickHouse query evaluating a core unit
-// over the [startMs, endMs] / stepMs evaluation grid: per-series grids via a
-// timeSeries*ToGrid aggregate (or a windowed aggregation for *_over_time),
-// then spatial aggregation with -ForEach combinators grouped by a canonical
-// JSON key of the projected label pairs.
-//
-// The heavy level is shaped to run on the shards: the top-level FROM is the
-// distributed samples table and the group-key join partner is a subquery on
-// the shard-local time series table, so the shard rewrite executes the join
-// and the per-(fingerprint, gkey) grid aggregation next to the data —
-// complete by fingerprint co-locality (see localTimeSeriesTable) — and the
-// initiator only merges per-series grid states and applies the spatial
-// -ForEach step. Same layout as the telemetrymetrics statement builder. The
-// windowed *_over_time form is the exception: its ARRAY JOIN level
-// aggregates on the shards the same way, but the group-key join happens at
-// the initiator over the already-reduced per-(series, index) rows — pushing
-// it down would not move any data off the initiator (the reduced rows arrive
-// there either way), so the combined ARRAY JOIN + JOIN form buys nothing.
-//
-// inlineFingerprints carries the matched set when it fits the inline limit;
-// nil means over the limit, where the group-key join restricts on its own
-// (the windowed form, whose fan-out query has no join, falls back to a
-// shard-local semi-join so it does not expand every series of the metric).
-//
-// The selector's data window is offset-shifted; the resulting grid indices
-// map 1:1 onto the query grid (output ts = startMs + i*stepMs). Grid
-// parameters are rendered as literals — they are aggregate-function
-// parameters, not bindable values.
-//
-// Statements nest builder-rendered SQL as text, so the returned args must be
-// ordered by where each fragment lands in the final statement: ClickHouse
-// binds ? placeholders by position. A JOIN renders before WHERE, so a joined
-// subquery's args precede the outer query's own condition args.
-//
-// Row shape: (gkey String, grid Array(Nullable(Float64))). gkey is
-// toJSONString of the sorted projected [key, value] pairs; NULL grid points
-// are absent points (the engine's "no value here"), which the -ForEach
-// combinators preserve: an index where every series is NULL aggregates to
-// NULL, and countForEach's 0 is mapped back to NULL.
-func buildUnitSQL(unit *coreUnit, metricNames []string, inlineFingerprints []uint64, dataStart, dataEnd int64, startMs, endMs, stepMs, lookbackMs int64) (string, []any, error) {
-	selStart := startMs - unit.offsetMs
-	selEnd := endMs - unit.offsetMs
-	stepSec := stepMs / 1000
-	if stepSec == 0 {
-		// Instant query: start == end, so the grid has one point for any
-		// positive step.
-		stepSec = 1
-	}
-	windowMs := unit.rangeMs
-	if unit.kind == unitInstant {
-		windowMs = lookbackMs
-	}
-	windowSec := windowMs / 1000
-
-	adjustedTsStart, tsTable := timeSeriesTableFor(dataStart, dataEnd)
-
-	// seriesSub computes fingerprint -> group key. It reads the local series
-	// table when it rides inside the shard-rewritten samples query, and the
-	// distributed one when it joins at the initiator (windowed form).
-	seriesSub := func(table string) (string, []any, error) {
-		sub := sqlbuilder.NewSelectBuilder()
-		sub.Select("fingerprint", groupKeyExpr(unit)+" AS gkey")
-		sub.From(fmt.Sprintf("%s.%s", databaseName, table))
-		if err := applySeriesConditions(sub, adjustedTsStart, dataEnd, unit.matchers); err != nil {
-			return "", nil, err
-		}
-		sub.GroupBy("fingerprint", "gkey")
-		q, args := sub.BuildWithFlavor(sqlbuilder.ClickHouse)
-		return q, args, nil
-	}
-
-	// samplesConditions adds the samples-side WHERE. The samples table is
-	// aliased "points" in every kind: under the group-key join both sides
-	// carry a fingerprint column, so the filter must qualify it. A nil
-	// inline set adds no fingerprint condition — the join restricts.
-	samplesConditions := func(sb *sqlbuilder.SelectBuilder, excludeStale bool) {
-		switch len(metricNames) {
-		case 0:
-			// No name constraint derivable; correct but unable to use the
-			// metric_name primary-key prefix.
-		case 1:
-			sb.Where(sb.EQ("metric_name", metricNames[0]))
-		default:
-			sb.Where(sb.In("metric_name", sqlbuilder.List(metricNames)))
-		}
-		// temporality precedes metric_name in the samples primary key; the
-		// fingerprints already come from these temporalities, so this only
-		// helps granule pruning.
-		sb.Where("temporality IN ['Cumulative', 'Unspecified']")
-		if inlineFingerprints != nil {
-			sb.Where("points.fingerprint " + inlineFingerprintFilter(inlineFingerprints))
-		}
-		// Left-open window: a sample exactly at the window's lower boundary
-		// is never used (range selectors and lookback are both left-open).
-		sb.Where(sb.GT("unix_milli", selStart-windowMs), sb.LTE("unix_milli", selEnd))
-		if excludeStale {
-			// PromQL excludes stale markers from range vectors. Instant
-			// selectors need the stale rows for shadowing instead.
-			sb.Where("bitAnd(flags, 1) = 0")
-		}
-	}
-
-	// joinedInner builds the shard-side SELECT for the single-pass kinds:
-	// grid expression per (fingerprint, gkey), group-key join against the
-	// local series table.
-	joinedInner := func(gridExpr string, excludeStale bool) (string, []any, error) {
-		seriesSQL, seriesArgs, err := seriesSub(localTimeSeriesTable(tsTable))
-		if err != nil {
-			return "", nil, err
-		}
-		sb := sqlbuilder.NewSelectBuilder()
-		sb.Select("series.gkey AS gkey", gridExpr+" AS grid")
-		sb.From(fmt.Sprintf("%s.%s AS points", databaseName, distributedSamplesV4))
-		sb.JoinWithOption(sqlbuilder.InnerJoin, fmt.Sprintf("(%s) AS series", seriesSQL), "points.fingerprint = series.fingerprint")
-		samplesConditions(sb, excludeStale)
-		sb.GroupBy("points.fingerprint", "series.gkey")
-		q, args := sb.BuildWithFlavor(sqlbuilder.ClickHouse)
-		// The join text renders before WHERE: its args come first.
-		return q, append(seriesArgs, args...), nil
-	}
-
-	var inner string
-	var innerArgs []any
-	var err error
-	switch unit.kind {
-	case unitInstant:
-		// Instant selection with stale shadowing: the grid value is the last
-		// non-stale sample in (t-lookback, t], absent when the overall last
-		// sample in that window is a stale marker (verified semantics: the
-		// -If combinator applies to the grid aggregates, and NULL comparisons
-		// make a stale-latest point absent).
-		gridParams := fmt.Sprintf("(fromUnixTimestamp64Milli(%d), fromUnixTimestamp64Milli(%d), %d, %d)", selStart, selEnd, stepSec, windowSec)
-		gridExpr := fmt.Sprintf(
-			"arrayMap((tall, tok, vok) -> if(tall IS NULL OR tok IS NULL OR tall != tok, NULL, vok), timeSeriesLastToGrid%s(fromUnixTimestamp64Milli(unix_milli), toFloat64(unix_milli)), timeSeriesLastToGridIf%s(fromUnixTimestamp64Milli(unix_milli), toFloat64(unix_milli), bitAnd(flags, 1) = 0), timeSeriesLastToGridIf%s(fromUnixTimestamp64Milli(unix_milli), value, bitAnd(flags, 1) = 0))",
-			gridParams, gridParams, gridParams,
-		)
-		inner, innerArgs, err = joinedInner(gridExpr, false)
-	case unitOverTime:
-		if unit.overFn == "last" {
-			// last_over_time == last non-stale sample in the window: the
-			// stale rows are already excluded in WHERE.
-			gridExpr := fmt.Sprintf(
-				"timeSeriesLastToGrid(fromUnixTimestamp64Milli(%d), fromUnixTimestamp64Milli(%d), %d, %d)(fromUnixTimestamp64Milli(unix_milli), value)",
-				selStart, selEnd, stepSec, windowSec,
-			)
-			inner, innerArgs, err = joinedInner(gridExpr, true)
-			break
-		}
-		inner, innerArgs, err = windowedInner(unit, samplesConditions, seriesSub, inlineFingerprints == nil, adjustedTsStart, dataEnd, tsTable, selStart, selEnd, stepMs, windowMs)
-	default: // unitRange
-		gridExpr := fmt.Sprintf(
-			"%s(fromUnixTimestamp64Milli(%d), fromUnixTimestamp64Milli(%d), %d, %d)(fromUnixTimestamp64Milli(unix_milli), value)",
-			gridFunction[unit.fn], selStart, selEnd, stepSec, windowSec,
-		)
-		if unit.fn == fnIncrease {
-			// increase == rate * range-seconds, exactly: extrapolatedRate
-			// divides by the range only when isRate.
-			gridExpr = fmt.Sprintf("arrayMap(x -> x * %d, %s)", windowSec, gridExpr)
-		}
-		inner, innerArgs, err = joinedInner(gridExpr, true)
-	}
-	if err != nil {
-		return "", nil, err
-	}
-
-	spatial := "maxForEach(grid)"
-	switch {
-	case !unit.hasAgg:
-		// Per-series output: one row per (labels-minus-__name__) group.
-		// Distinct fingerprints can collapse onto the same projected label
-		// set only via a regex __name__ selector over metrics with identical
-		// other labels; maxForEach is a deterministic NULL-skipping merge and
-		// the identity for the overwhelmingly common one-fingerprint group.
-	case unit.aggOp.String() == "count":
-		// count over an all-absent index is an absent point, not 0.
-		spatial = "arrayMap(c -> if(c = 0, NULL, toFloat64(c)), countForEach(grid))"
-	default:
-		spatial = fmt.Sprintf("%s(grid)", aggForEach[unit.aggOp.String()])
-	}
-
-	query := fmt.Sprintf("SELECT gkey, %s AS grid FROM (%s) GROUP BY gkey %s", spatial, inner, gridFunctionsSetting)
-	return query, innerArgs, nil
-}
-
-// windowedInner builds the avg/min/max/sum/count _over_time form: each
-// sample fans out to every grid index k whose window (t_k - range, t_k]
-// contains it (ARRAY JOIN), aggregates per (fingerprint, k) — shard-side
-// partials over the distributed table — then assembles the positional grid
-// and joins the group key at the initiator over the reduced rows. The
-// group-key subquery reads the distributed series table here because it does
-// not ride inside a shard-rewritten query.
-//
-// This is the one form whose samples query has no series join, so an
-// over-the-limit fingerprint set (semiJoin) must fall back to the
-// shard-local semi-join: without it the fan-out would expand every series of
-// the metric and discard the unmatched ones only at the group-key join.
-func windowedInner(unit *coreUnit, samplesConditions func(*sqlbuilder.SelectBuilder, bool), seriesSub func(string) (string, []any, error), semiJoin bool, adjustedTsStart, dataEnd int64, tsTable string, selStart, selEnd, stepMs, windowMs int64) (string, []any, error) {
-	aggExpr := map[string]string{
-		"avg":   "avg(value)",
-		"min":   "min(value)",
-		"max":   "max(value)",
-		"sum":   "sum(value)",
-		"count": "toFloat64(count(value))",
-	}[unit.overFn]
-	effStepMs := stepMs
-	if effStepMs == 0 {
-		effStepMs = 1000
-	}
-	lastIdx := (selEnd - selStart) / effStepMs
-
-	perWindow := sqlbuilder.NewSelectBuilder()
-	perWindow.Select("fingerprint", "k", aggExpr+" AS v")
-	perWindow.From(fmt.Sprintf(
-		"%s.%s AS points ARRAY JOIN range(toUInt64(greatest(0, intDiv(unix_milli - %d + %d - 1, %d))), toUInt64(least(%d, intDiv(unix_milli + %d - 1 - %d, %d)) + 1)) AS k",
-		databaseName, distributedSamplesV4,
-		selStart, effStepMs, effStepMs,
-		lastIdx, windowMs, selStart, effStepMs,
-	))
-	samplesConditions(perWindow, true)
-	if semiJoin {
-		sub := sqlbuilder.NewSelectBuilder()
-		sub.Select("fingerprint")
-		sub.From(fmt.Sprintf("%s.%s", databaseName, localTimeSeriesTable(tsTable)))
-		if err := applySeriesConditions(sub, adjustedTsStart, dataEnd, unit.matchers); err != nil {
-			return "", nil, err
-		}
-		perWindow.Where(perWindow.In("points.fingerprint", sub))
-	}
-	perWindow.GroupBy("fingerprint", "k")
-	perWindowSQL, perWindowArgs := perWindow.BuildWithFlavor(sqlbuilder.ClickHouse)
-
-	grids := fmt.Sprintf(
-		"SELECT fingerprint, arrayMap(i -> if(indexOf(ks, i) = 0, NULL, vs[indexOf(ks, i)]), range(toUInt64(%d))) AS grid FROM (SELECT fingerprint, groupArray(k) AS ks, groupArray(v) AS vs FROM (%s) GROUP BY fingerprint)",
-		lastIdx+1, perWindowSQL,
-	)
-
-	seriesSQL, seriesArgs, err := seriesSub(tsTable)
-	if err != nil {
-		return "", nil, err
-	}
-	inner := fmt.Sprintf(
-		"SELECT series.gkey AS gkey, points.grid AS grid FROM (%s) AS points INNER JOIN (%s) AS series ON points.fingerprint = series.fingerprint",
-		grids, seriesSQL,
-	)
-	return inner, append(perWindowArgs, seriesArgs...), nil
-}
-
-// groupKeyExpr renders the canonical group key for a unit: the sorted
-// [key, value] pairs of the projected labels, JSON-encoded.
-//   - by (a, b): keep only the listed labels (absent labels stay absent,
-//     matching PromQL's by() over missing labels);
-//   - without (a, b): keep everything except the listed labels and __name__;
-//   - no aggregation: keep everything including __name__ — even when the
-//     unit drops the name from its OUTPUT, the key must keep it so distinct
-//     metrics never merge in SQL; executeUnit strips the name afterwards and
-//     turns a post-strip collision into the engine's duplicate-labelset
-//     error instead of a silently invented merge.
-func groupKeyExpr(unit *coreUnit) string {
-	// An empty label value means "label absent" in Prometheus; the stored
-	// labels JSON can carry empty attribute values, which must not become
-	// output labels or group keys.
-	pairs := "arraySort(JSONExtractKeysAndValues(labels, 'String'))"
-	if !unit.hasAgg {
-		return fmt.Sprintf("toJSONString(arrayFilter(p -> p.2 != '', %s))", pairs)
-	}
-	if unit.by {
-		if len(unit.grouping) == 0 {
-			return "'[]'"
-		}
-		return fmt.Sprintf("toJSONString(arrayFilter(p -> p.2 != '' AND p.1 IN (%s), %s))", quotedList(unit.grouping), pairs)
-	}
-	excluded := append([]string{metricNameLabel}, unit.grouping...)
-	return fmt.Sprintf("toJSONString(arrayFilter(p -> p.2 != '' AND p.1 NOT IN (%s), %s))", quotedList(excluded), pairs)
-}
-
-func quotedList(items []string) string {
-	quoted := make([]string, len(items))
-	for i, s := range items {
-		quoted[i] = "'" + strings.ReplaceAll(s, "'", "\\'") + "'"
-	}
-	return strings.Join(quoted, ", ")
-}

pkg/prometheus/clickhouseprometheusv2/transpiler_test.go

@@ -1,517 +0,0 @@
-package clickhouseprometheusv2
-
-import (
-	"context"
-	"sync/atomic"
-	"testing"
-	"time"
-
-	cmock "github.com/SigNoz/clickhouse-go-mock"
-	"github.com/SigNoz/signoz/pkg/errors"
-	"github.com/SigNoz/signoz/pkg/prometheus"
-	"github.com/prometheus/prometheus/model/labels"
-	"github.com/prometheus/prometheus/promql/parser"
-	"github.com/stretchr/testify/assert"
-	"github.com/stretchr/testify/require"
-)
-
-func parse(t *testing.T, q string) parser.Expr {
-	t.Helper()
-	expr, err := parser.NewParser(parser.Options{}).ParseExpr(q)
-	require.NoError(t, err)
-	return expr
-}
-
-func TestClassifyFullShapes(t *testing.T) {
-	tests := []struct {
-		name  string
-		query string
-		check func(t *testing.T, u *coreUnit)
-	}{
-		{
-			name:  "sum by rate",
-			query: `sum by (pod) (rate(http_requests_total{job="api"}[5m]))`,
-			check: func(t *testing.T, u *coreUnit) {
-				assert.Equal(t, fnRate, u.fn)
-				assert.Equal(t, int64(300_000), u.rangeMs)
-				assert.True(t, u.hasAgg)
-				assert.True(t, u.by)
-				assert.Equal(t, []string{"pod"}, u.grouping)
-			},
-		},
-		{
-			name:  "bare increase with offset",
-			query: `increase(errors_total[10m] offset 30m)`,
-			check: func(t *testing.T, u *coreUnit) {
-				assert.Equal(t, fnIncrease, u.fn)
-				assert.Equal(t, int64(1_800_000), u.offsetMs)
-				assert.False(t, u.hasAgg)
-			},
-		},
-		{
-			name:  "avg without over delta",
-			query: `avg without (instance) (delta(gauge_metric[15m]))`,
-			check: func(t *testing.T, u *coreUnit) {
-				assert.Equal(t, fnDelta, u.fn)
-				assert.True(t, u.hasAgg)
-				assert.False(t, u.by)
-			},
-		},
-		{
-			name:  "scalar pipeline with comparison",
-			query: `sum(rate(x[5m])) * 100 > 5`,
-			check: func(t *testing.T, u *coreUnit) {
-				require.Len(t, u.ops, 2)
-				assert.Equal(t, parser.ItemType(parser.MUL), u.ops[0].op)
-				assert.Equal(t, 100.0, u.ops[0].scalar)
-				assert.Equal(t, parser.ItemType(parser.GTR), u.ops[1].op)
-			},
-		},
-		{
-			name:  "scalar on left with unary minus",
-			query: `-1 * sum(rate(x[5m]))`,
-			check: func(t *testing.T, u *coreUnit) {
-				require.Len(t, u.ops, 1)
-				assert.True(t, u.ops[0].scalarOnLeft)
-				assert.Equal(t, -1.0, u.ops[0].scalar)
-			},
-		},
-		{
-			name:  "bool comparison",
-			query: `sum(rate(x[5m])) >= bool 0.5`,
-			check: func(t *testing.T, u *coreUnit) {
-				require.Len(t, u.ops, 1)
-				assert.True(t, u.ops[0].returnBool)
-			},
-		},
-		{
-			name:  "irate utf8 name",
-			query: `sum by ("k8s.pod.name") (irate({"k8s.container.cpu.time"}[2m]))`,
-			check: func(t *testing.T, u *coreUnit) {
-				assert.Equal(t, fnIRate, u.fn)
-				assert.Equal(t, []string{"k8s.pod.name"}, u.grouping)
-			},
-		},
-		{
-			name:  "bare instant selector keeps name",
-			query: `up{job="api"}`,
-			check: func(t *testing.T, u *coreUnit) {
-				assert.Equal(t, unitInstant, u.kind)
-				assert.True(t, u.keepsName())
-			},
-		},
-		{
-			name:  "gauge aggregation",
-			query: `sum by (pod) (container_memory offset 5m)`,
-			check: func(t *testing.T, u *coreUnit) {
-				assert.Equal(t, unitInstant, u.kind)
-				assert.Equal(t, int64(300_000), u.offsetMs)
-				assert.True(t, u.hasAgg)
-				assert.False(t, u.keepsName())
-			},
-		},
-		{
-			name:  "gauge comparison keeps name",
-			query: `container_memory > 100`,
-			check: func(t *testing.T, u *coreUnit) {
-				assert.Equal(t, unitInstant, u.kind)
-				assert.True(t, u.keepsName())
-			},
-		},
-		{
-			name:  "gauge arithmetic drops name",
-			query: `container_memory / 1024`,
-			check: func(t *testing.T, u *coreUnit) {
-				assert.Equal(t, unitInstant, u.kind)
-				assert.False(t, u.keepsName())
-			},
-		},
-		{
-			name:  "avg_over_time",
-			query: `max by (node) (avg_over_time(load1[10m]))`,
-			check: func(t *testing.T, u *coreUnit) {
-				assert.Equal(t, unitOverTime, u.kind)
-				assert.Equal(t, "avg", u.overFn)
-				assert.Equal(t, int64(600_000), u.rangeMs)
-			},
-		},
-		{
-			name:  "last_over_time keeps name",
-			query: `last_over_time(load1[10m])`,
-			check: func(t *testing.T, u *coreUnit) {
-				assert.Equal(t, unitOverTime, u.kind)
-				assert.Equal(t, "last", u.overFn)
-				assert.True(t, u.keepsName())
-			},
-		},
-	}
-	for _, tt := range tests {
-		t.Run(tt.name, func(t *testing.T) {
-			plan, ok := classify(parse(t, tt.query), testGrid(60_000))
-			require.True(t, ok, "expected transpilable")
-			require.True(t, plan.full, "expected full compilation")
-			require.Len(t, plan.units, 1)
-			tt.check(t, &plan.units[0].core)
-		})
-	}
-}
-
-func TestClassifyFallbackShapes(t *testing.T) {
-	queries := []struct {
-		name  string
-		query string
-		step  int64
-	}{
-		{"default-resolution subquery", `max_over_time(rate(x[5m])[30m:])`, 60_000},
-		{"at modifier", `sum(rate(x[5m] @ 1609746000))`, 60_000},
-		{"at modifier on gauge", `sum(container_memory @ 1609746000)`, 60_000},
-		{"sub-second step", `sum(rate(x[5m]))`, 500},
-		{"sub-second range", `sum(rate(x[1500ms]))`, 60_000},
-		{"by __name__ full", `sum by (__name__) (rate({__name__=~"a|b"}[5m]))`, 60_000},
-		{"quantile_over_time unsupported", `quantile_over_time(0.9, load1[10m])`, 60_000},
-	}
-	for _, tt := range queries {
-		t.Run(tt.name, func(t *testing.T) {
-			_, ok := classify(parse(t, tt.query), testGrid(tt.step))
-			assert.False(t, ok, "expected fallback for %s", tt.query)
-		})
-	}
-}
-
-func TestClassifyHybridShapes(t *testing.T) {
-	tests := []struct {
-		name          string
-		query         string
-		wantUnits     int
-		wantRewritten string
-	}{
-		{
-			name:          "histogram quantile",
-			query:         `histogram_quantile(0.95, sum by (le) (rate(http_bucket[5m])))`,
-			wantUnits:     1,
-			wantRewritten: `histogram_quantile(0.95, __signoz_transpiled_0__)`,
-		},
-		{
-			name:          "topk over compiled",
-			query:         `topk(5, sum by (pod) (rate(x[5m])))`,
-			wantUnits:     1,
-			wantRewritten: `topk(5, __signoz_transpiled_0__)`,
-		},
-		{
-			name:          "ratio of compiled units",
-			query:         `sum(rate(a[5m])) / sum(rate(b[5m]))`,
-			wantUnits:     2,
-			wantRewritten: `__signoz_transpiled_0__ / __signoz_transpiled_1__`,
-		},
-		{
-			name:          "or vector zero",
-			query:         `sum(rate(a[5m])) or vector(0)`,
-			wantUnits:     1,
-			wantRewritten: `__signoz_transpiled_0__ or vector(0)`,
-		},
-		{
-			name:          "quantile agg over compiled rate",
-			query:         `quantile(0.9, rate(x[5m]))`,
-			wantUnits:     1,
-			wantRewritten: `quantile(0.9, __signoz_transpiled_0__)`,
-		},
-		{
-			name:          "non-literal scalar side stays engine-side",
-			query:         `sum(rate(x[5m])) * scalar(y)`,
-			wantUnits:     1,
-			wantRewritten: `__signoz_transpiled_0__ * scalar(y)`,
-		},
-		{
-			name:          "compiled mixed with raw selector",
-			query:         `sum by (pod) (rate(a[5m])) / on (pod) group_left () b`,
-			wantUnits:     1,
-			wantRewritten: `__signoz_transpiled_0__ / on (pod) group_left () b`,
-		},
-	}
-	for _, tt := range tests {
-		t.Run(tt.name, func(t *testing.T) {
-			plan, ok := classify(parse(t, tt.query), testGrid(60_000))
-			require.True(t, ok)
-			assert.False(t, plan.full)
-			assert.Len(t, plan.units, tt.wantUnits)
-			assert.Equal(t, tt.wantRewritten, plan.rewritten)
-		})
-	}
-}
-
-func TestClassifyHybridGuards(t *testing.T) {
-	t.Run("no substitution under on(__name__)", func(t *testing.T) {
-		plan, ok := classify(parse(t, `sum(rate(a[5m])) * on (__name__) b`), testGrid(60_000))
-		_ = plan
-		assert.False(t, ok, "matching on __name__ must not see synthetic names")
-	})
-	t.Run("no substitution inside @-pinned subquery", func(t *testing.T) {
-		_, ok := classify(parse(t, `max_over_time(rate(x[5m])[30m:1m] @ 1609746000)`), testGrid(60_000))
-		assert.False(t, ok)
-	})
-}
-
-// The alert-smoothing idiom: units inside a fixed-resolution subquery
-// evaluate on the subquery grid — epoch-aligned multiples of the resolution,
-// starting strictly after (outer start - range), exactly as the engine
-// derives it.
-func TestClassifySubqueryUnits(t *testing.T) {
-	grid := gridContext{startMs: 1_700_000_030_000, endMs: 1_700_007_200_000, stepMs: 60_000}
-
-	plan, ok := classify(parse(t, `min_over_time((sum by (ns) (increase(x[5m])))[10m:5m]) > 0`), grid)
-	require.True(t, ok)
-	require.False(t, plan.full)
-	require.Len(t, plan.units, 1)
-	assert.Equal(t, `min_over_time(__signoz_transpiled_0__[10m:5m]) > 0`, plan.rewritten)
-
-	unit := plan.units[0]
-	// lower bound = outer start - range = 1_699_999_430_000; first multiple
-	// of 300_000 strictly greater is 1_699_999_500_000.
-	assert.Equal(t, int64(1_699_999_500_000), unit.grid.startMs)
-	assert.Equal(t, grid.endMs, unit.grid.endMs)
-	assert.Equal(t, int64(300_000), unit.grid.stepMs)
-	assert.Equal(t, fnIncrease, unit.core.fn)
-
-	t.Run("subquery offset shifts the grid", func(t *testing.T) {
-		plan, ok := classify(parse(t, `max_over_time((sum(rate(x[5m])))[10m:5m] offset 30m)`), grid)
-		require.True(t, ok)
-		require.Len(t, plan.units, 1)
-		// lower = start - offset - range = 1_699_997_630_000 -> first
-		// multiple of 300_000 above = 1_699_997_700_000; end shifts too.
-		assert.Equal(t, int64(1_699_997_700_000), plan.units[0].grid.startMs)
-		assert.Equal(t, grid.endMs-1_800_000, plan.units[0].grid.endMs)
-	})
-
-	t.Run("mollusk ratio-inside-subquery idiom", func(t *testing.T) {
-		q := `min_over_time(((sum by (a) (rate(m1[5m]))) / (avg by (a) (m2)))[5m:1m])`
-		plan, ok := classify(parse(t, q), grid)
-		require.True(t, ok)
-		// Both sides compile on the subquery grid: the rate side and the
-		// gauge aggregation side; the engine joins them and smooths.
-		require.Len(t, plan.units, 2)
-		assert.Equal(t, int64(60_000), plan.units[0].grid.stepMs)
-		assert.Equal(t, unitInstant, plan.units[1].core.kind)
-		assert.Contains(t, plan.rewritten, `__signoz_transpiled_0__ / __signoz_transpiled_1__`)
-	})
-}
-
-func TestBuildUnitSQL(t *testing.T) {
-	unit := &coreUnit{
-		fn:       fnRate,
-		rangeMs:  300_000,
-		hasAgg:   true,
-		aggOp:    parser.SUM,
-		by:       true,
-		grouping: []string{"pod"},
-		matchers: []*labels.Matcher{mustMatcher(t, labels.MatchEqual, "__name__", "http_requests_total")},
-	}
-	sql, args, err := buildUnitSQL(unit, []string{"http_requests_total"}, []uint64{7, 42}, 1_699_999_700_000, 1_700_003_600_000, 1_700_000_000_000, 1_700_003_600_000, 60_000, 300_000)
-	require.NoError(t, err)
-
-	assert.Contains(t, sql, "timeSeriesRateToGrid(fromUnixTimestamp64Milli(1700000000000), fromUnixTimestamp64Milli(1700003600000), 60, 300)(fromUnixTimestamp64Milli(unix_milli), value)")
-	assert.Contains(t, sql, "unix_milli > ? AND unix_milli <= ?")
-	assert.Contains(t, sql, "bitAnd(flags, 1) = 0")
-	assert.Contains(t, sql, "sumForEach(grid)")
-	// The group-key join rides inside the shard query: distributed samples
-	// at the top level, the local series table in the join subquery, the
-	// grid aggregation grouped per (fingerprint, gkey) shard-side.
-	assert.Contains(t, sql, "FROM signoz_metrics.distributed_samples_v4 AS points INNER JOIN (SELECT fingerprint,")
-	assert.Contains(t, sql, "FROM signoz_metrics.time_series_v4 WHERE")
-	assert.Contains(t, sql, "GROUP BY points.fingerprint, series.gkey")
-	assert.Contains(t, sql, "points.fingerprint IN (7, 42)")
-	assert.Contains(t, sql, `toJSONString(arrayFilter(p -> p.2 != '' AND p.1 IN ('pod'),`)
-	assert.Contains(t, sql, "SETTINGS allow_experimental_ts_to_grid_aggregate_function = 1")
-	// Args follow placeholder order: the joined series subquery renders
-	// before the samples WHERE.
-	assert.Equal(t, []any{"http_requests_total", int64(1_699_999_200_000), int64(1_700_003_600_000), "http_requests_total", int64(1_699_999_700_000), int64(1_700_003_600_000)}, args)
-}
-
-func TestBuildUnitSQLIncreaseAndOffset(t *testing.T) {
-	unit := &coreUnit{
-		fn:       fnIncrease,
-		rangeMs:  600_000,
-		offsetMs: 1_800_000,
-		matchers: []*labels.Matcher{mustMatcher(t, labels.MatchEqual, "__name__", "errors_total")},
-	}
-	sql, _, err := buildUnitSQL(unit, nil, []uint64{7}, 1_699_997_600_000, 1_700_001_800_000, 1_700_000_000_000, 1_700_003_600_000, 60_000, 300_000)
-	require.NoError(t, err)
-
-	// Grid and window shift by the offset; increase multiplies rate by the
-	// range in seconds.
-	assert.Contains(t, sql, "fromUnixTimestamp64Milli(1699998200000), fromUnixTimestamp64Milli(1700001800000)")
-	assert.Contains(t, sql, "arrayMap(x -> x * 600, timeSeriesRateToGrid")
-	assert.Contains(t, sql, "maxForEach(grid)")
-}
-
-func TestBuildUnitSQLOverLimitJoinOnly(t *testing.T) {
-	// Past the inline limit no fingerprint filter is rendered: the series
-	// join restricts to the matched fingerprints on its own.
-	unit := &coreUnit{
-		fn:       fnRate,
-		rangeMs:  300_000,
-		hasAgg:   true,
-		aggOp:    parser.SUM,
-		by:       true,
-		matchers: []*labels.Matcher{mustMatcher(t, labels.MatchEqual, "__name__", "http_requests_total")},
-	}
-	sql, _, err := buildUnitSQL(unit, []string{"http_requests_total"}, nil, 1_699_999_700_000, 1_700_003_600_000, 1_700_000_000_000, 1_700_003_600_000, 60_000, 300_000)
-	require.NoError(t, err)
-
-	assert.NotContains(t, sql, "points.fingerprint IN")
-	assert.Contains(t, sql, "INNER JOIN (SELECT fingerprint,")
-	assert.Contains(t, sql, "FROM signoz_metrics.time_series_v4 WHERE")
-}
-
-func TestBuildUnitSQLOverLimitWindowedSemiJoin(t *testing.T) {
-	// The windowed *_over_time fan-out has no series join, so the over-limit
-	// regime falls back to the shard-local semi-join instead of expanding
-	// every series of the metric.
-	unit := &coreUnit{
-		kind:     unitOverTime,
-		overFn:   "avg",
-		rangeMs:  600_000,
-		matchers: []*labels.Matcher{mustMatcher(t, labels.MatchEqual, "__name__", "node_load1")},
-	}
-	sql, _, err := buildUnitSQL(unit, []string{"node_load1"}, nil, 1_699_999_400_000, 1_700_003_600_000, 1_700_000_000_000, 1_700_003_600_000, 60_000, 300_000)
-	require.NoError(t, err)
-
-	assert.Contains(t, sql, "points.fingerprint IN (SELECT fingerprint FROM signoz_metrics.time_series_v4 WHERE ")
-	assert.Contains(t, sql, "ARRAY JOIN range(")
-}
-
-func TestApplyScalarOps(t *testing.T) {
-	f := func(v float64) *float64 { return &v }
-
-	t.Run("arithmetic chain", func(t *testing.T) {
-		values := []*float64{f(2), nil, f(4)}
-		applyScalarOps([]scalarOp{{op: parser.MUL, scalar: 100}, {op: parser.ADD, scalar: 1}}, values)
-		require.NotNil(t, values[0])
-		assert.Equal(t, 201.0, *values[0])
-		assert.Nil(t, values[1])
-		assert.Equal(t, 401.0, *values[2])
-	})
-
-	t.Run("comparison filters points", func(t *testing.T) {
-		values := []*float64{f(1), f(10)}
-		applyScalarOps([]scalarOp{{op: parser.GTR, scalar: 5}}, values)
-		assert.Nil(t, values[0])
-		require.NotNil(t, values[1])
-		assert.Equal(t, 10.0, *values[1], "filter comparisons keep the original value")
-	})
-
-	t.Run("bool comparison emits 0/1", func(t *testing.T) {
-		values := []*float64{f(1), f(10)}
-		applyScalarOps([]scalarOp{{op: parser.GTR, scalar: 5, returnBool: true}}, values)
-		assert.Equal(t, 0.0, *values[0])
-		assert.Equal(t, 1.0, *values[1])
-	})
-
-	t.Run("scalar on left division", func(t *testing.T) {
-		values := []*float64{f(4)}
-		applyScalarOps([]scalarOp{{op: parser.DIV, scalar: 100, scalarOnLeft: true}}, values)
-		assert.Equal(t, 25.0, *values[0])
-	})
-}
-
-func TestLabelsFromGroupKey(t *testing.T) {
-	lset, err := labelsFromGroupKey(`[["pod","api-0"],["ns","prod"]]`)
-	require.NoError(t, err)
-	assert.Equal(t, "api-0", lset.Get("pod"))
-	assert.Equal(t, "prod", lset.Get("ns"))
-
-	empty, err := labelsFromGroupKey(`[]`)
-	require.NoError(t, err)
-	assert.True(t, empty.IsEmpty())
-}
-
-// testGrid is a 2h query grid ending on a round timestamp.
-func testGrid(stepMs int64) gridContext {
-	return gridContext{startMs: 1_700_000_000_000, endMs: 1_700_007_200_000, stepMs: stepMs}
-}
-
-// A bool comparison returns 0/1, not the sample, so the engine drops
-// __name__; keeping it would change downstream vector matching.
-func TestKeepsName_BoolComparisonDropsName(t *testing.T) {
-	plan, ok := classify(parse(t, `up > bool 0`), testGrid(60_000))
-	require.True(t, ok)
-	assert.False(t, plan.units[0].core.keepsName())
-
-	plan, ok = classify(parse(t, `up > 0`), testGrid(60_000))
-	require.True(t, ok)
-	assert.True(t, plan.units[0].core.keepsName())
-}
-
-// timeSeriesLastToGrid widens its window to max(window, step) — probed on
-// 25.12 — so Last-style units at window < step must fall back or they would
-// resurrect samples the engine's lookback already dropped.
-func TestTryExecuteRange_LastStyleWindowBelowStepFallsBack(t *testing.T) {
-	c, _ := newTestClient(t, prometheus.ClickhouseV2Config{})
-	e := &executor{client: c, parser: prometheus.NewParser()}
-
-	start := time.UnixMilli(1_700_000_000_000)
-	end := time.UnixMilli(1_700_003_600_000)
-
-	_, ok, err := e.TryExecuteRange(context.Background(), `sum by (pod) (up)`, start, end, time.Hour)
-	require.NoError(t, err)
-	assert.False(t, ok, "instant selection at step > lookback must not transpile")
-
-	_, ok, err = e.TryExecuteRange(context.Background(), `last_over_time(up[10m])`, start, end, time.Hour)
-	require.NoError(t, err)
-	assert.False(t, ok, "last_over_time at range < step must not transpile")
-}
-
-// Transpiled results never pass the engine's sample limiter, so the grid
-// cells (series x grid width) must be budgeted before the arrays exist —
-// otherwise a wide query rebuilds the OOM this provider exists to prevent.
-func TestExecuteUnit_GridCellBudget(t *testing.T) {
-	c, store := newTestClient(t, prometheus.ClickhouseV2Config{MaxFetchedSamples: 100})
-	e := &executor{client: c, parser: prometheus.NewParser()}
-
-	store.Mock().ExpectQuery("SELECT fingerprint, any\\(labels\\)").WithArgs("up", int64(1_699_999_200_000), int64(1_700_003_600_000)).WillReturnRows(cmock.NewRows(seriesCols, [][]any{
-		{uint64(1), `{"__name__":"up","instance":"a"}`},
-		{uint64(2), `{"__name__":"up","instance":"b"}`},
-	}))
-
-	plan, ok := classify(parse(t, `sum(rate(up[5m]))`), gridContext{startMs: 1_700_000_000_000, endMs: 1_700_003_600_000, stepMs: 60_000})
-	require.True(t, ok)
-
-	var cells atomic.Int64
-	// 2 series x 61 grid points = 122 cells > 100.
-	_, err := e.executeUnit(context.Background(), &plan.units[0].core, plan.units[0].grid, &cells)
-	require.Error(t, err)
-	assert.True(t, errors.Ast(err, errors.TypeInvalidInput), "budget refusal must be typed invalid input, got %v", err)
-}
-
-// Two metrics collapsing onto one labelset after the name drop is the
-// engine's duplicate-labelset error; silently merging them would invent a
-// series no engine would produce.
-func TestExecuteUnit_NameCollisionErrors(t *testing.T) {
-	c, store := newTestClient(t, prometheus.ClickhouseV2Config{})
-	e := &executor{client: c, parser: prometheus.NewParser()}
-
-	store.Mock().ExpectQuery("SELECT fingerprint, any\\(labels\\)").WithArgs("^(?:a|b)$", int64(1_699_999_200_000), int64(1_700_003_600_000)).WillReturnRows(cmock.NewRows(seriesCols, [][]any{
-		{uint64(1), `{"__name__":"a","job":"x"}`},
-		{uint64(2), `{"__name__":"b","job":"x"}`},
-	}))
-	store.Mock().ExpectQuery("SELECT gkey").
-		WithArgs("^(?:a|b)$", int64(1_699_999_200_000), int64(1_700_003_600_000), "a", "b", int64(1_699_999_700_000), int64(1_700_003_600_000)).
-		WillReturnRows(cmock.NewRows(gkeyCols, [][]any{
-			{`[["__name__","a"],["job","x"]]`, []*float64{f64(1)}},
-			{`[["__name__","b"],["job","x"]]`, []*float64{f64(2)}},
-		}))
-
-	plan, ok := classify(parse(t, `rate({__name__=~"a|b"}[5m])`), gridContext{startMs: 1_700_000_000_000, endMs: 1_700_003_600_000, stepMs: 60_000})
-	require.True(t, ok)
-
-	var cells atomic.Int64
-	_, err := e.executeUnit(context.Background(), &plan.units[0].core, plan.units[0].grid, &cells)
-	require.Error(t, err)
-	assert.Contains(t, err.Error(), "vector cannot contain metrics with the same labelset")
-}
-
-var gkeyCols = []cmock.ColumnType{
-	{Name: "gkey", Type: "String"},
-	{Name: "grid", Type: "Array(Nullable(Float64))"},
-}
-
-func f64(v float64) *float64 { return &v }

pkg/prometheus/config.go

@@ -13,11 +13,6 @@ type ActiveQueryTrackerConfig struct {
 	MaxConcurrent int    `mapstructure:"max_concurrent"`
 }
 
-type ClickhouseV2Config struct {
-	MaxFetchedSeries  int   `mapstructure:"max_fetched_series"`
-	MaxFetchedSamples int64 `mapstructure:"max_fetched_samples"`
-}
-
 type Config struct {
 	ActiveQueryTrackerConfig ActiveQueryTrackerConfig `mapstructure:"active_query_tracker"`
 
@@ -29,13 +24,6 @@ type Config struct {
 
 	// Timeout is the maximum time a query is allowed to run before being aborted.
 	Timeout time.Duration `mapstructure:"timeout"`
-
-	// ProviderName selects the storage provider: "clickhouse" (default) or
-	// "clickhousev2".
-	ProviderName string `mapstructure:"provider"`
-
-	// ClickhouseV2 configures the clickhousev2 provider.
-	ClickhouseV2 ClickhouseV2Config `mapstructure:"clickhousev2"`
 }
 
 func NewConfigFactory() factory.ConfigFactory {
@@ -49,12 +37,7 @@ func newConfig() factory.Config {
 			Path:          "",
 			MaxConcurrent: 20,
 		},
-		Timeout:      2 * time.Minute,
-		ProviderName: "clickhouse",
-		ClickhouseV2: ClickhouseV2Config{
-			MaxFetchedSeries:  500_000,
-			MaxFetchedSamples: 50_000_000,
-		},
+		Timeout: 2 * time.Minute,
 	}
 }
 
@@ -62,18 +45,9 @@ func (c Config) Validate() error {
 	if c.Timeout <= 0 {
 		return errors.Newf(errors.TypeInvalidInput, errors.CodeInvalidInput, "prometheus::timeout must be greater than 0")
 	}
-	if c.ProviderName != "" && c.ProviderName != "clickhouse" && c.ProviderName != "clickhousev2" {
-		return errors.Newf(errors.TypeInvalidInput, errors.CodeInvalidInput, "prometheus::provider must be one of [clickhouse, clickhousev2], got %q", c.ProviderName)
-	}
-	if c.ClickhouseV2.MaxFetchedSeries < 0 || c.ClickhouseV2.MaxFetchedSamples < 0 {
-		return errors.Newf(errors.TypeInvalidInput, errors.CodeInvalidInput, "prometheus::clickhousev2 limits must not be negative")
-	}
 	return nil
 }
 
 func (c Config) Provider() string {
-	if c.ProviderName == "" {
-		return "clickhouse"
-	}
-	return c.ProviderName
+	return "clickhouse"
 }

pkg/prometheus/prometheus.go

@@ -35,9 +35,3 @@ type StatementRecorder interface {
 type StatementCapturer interface {
 	CapturingStorage() (storage.Queryable, StatementRecorder)
 }
-
-// ProviderClickhouseV2 is the clickhousev2 provider name: the factory
-// registration, the prometheus::provider config value and the
-// X-SigNoz-PromQL-Provider request header all use it, so they cannot drift
-// apart.
-const ProviderClickhouseV2 = "clickhousev2"

pkg/prometheus/traits.go

@@ -1,49 +0,0 @@
-package prometheus
-
-import (
-	"context"
-
-	"github.com/prometheus/prometheus/promql/parser"
-)
-
-type queryTraitsKey struct{}
-
-// QueryTraits carries per-query facts a storage implementation cannot derive
-// from SelectHints alone. Call sites that parse the PromQL expression attach
-// traits to the context before handing it to the engine; storages treat a
-// missing traits value as "unknown" and stay conservative.
-type QueryTraits struct {
-	// SubqueryFree is true when the query contains no subquery expression.
-	// Subquery selectors are evaluated at the subquery's own step, but
-	// SelectHints.Step always carries the top-level step, so step-aligned
-	// storage optimizations (e.g. keeping only the last sample per step
-	// bucket) are safe only when this is true.
-	SubqueryFree bool
-}
-
-// DetectQueryTraits derives QueryTraits from a parsed PromQL expression.
-func DetectQueryTraits(expr parser.Expr) QueryTraits {
-	subqueryFree := true
-	parser.Inspect(expr, func(node parser.Node, _ []parser.Node) error {
-		if _, ok := node.(*parser.SubqueryExpr); ok {
-			subqueryFree = false
-		}
-		return nil
-	})
-	return QueryTraits{SubqueryFree: subqueryFree}
-}
-
-// NewContextWithQueryTraits returns a context carrying the given traits.
-func NewContextWithQueryTraits(ctx context.Context, traits QueryTraits) context.Context {
-	return context.WithValue(ctx, queryTraitsKey{}, traits)
-}
-
-// QueryTraitsFromContext returns the traits attached to ctx, if any.
-//
-// Context is used here, unlike for backend selection, because traits must
-// cross the promql engine to reach storage.Querier.Select, and the engine's
-// interfaces offer no other channel; the alternative is a Prometheus fork.
-func QueryTraitsFromContext(ctx context.Context) (QueryTraits, bool) {
-	traits, ok := ctx.Value(queryTraitsKey{}).(QueryTraits)
-	return traits, ok
-}

pkg/querier/api.go

@@ -50,7 +50,6 @@ func (handler *handler) QueryRange(rw http.ResponseWriter, req *http.Request) {
 		render.Error(rw, err)
 		return
 	}
-	queryRangeRequest.PromQLProvider = req.Header.Get("X-SigNoz-PromQL-Provider")
 
 	// Validate the query request
 	if err := queryRangeRequest.Validate(); err != nil {

pkg/querier/builder_query.go

@@ -17,6 +17,7 @@ import (
 	"github.com/SigNoz/signoz/pkg/types/instrumentationtypes"
 	qbtypes "github.com/SigNoz/signoz/pkg/types/querybuildertypes/querybuildertypesv5"
 	"github.com/SigNoz/signoz/pkg/types/telemetrytypes"
+	"github.com/SigNoz/signoz/pkg/valuer"
 )
 
 const traceOutsideRangeWarn = "Query %s references a trace_id that exists between %s and %s (UTC) but lies outside the selected time range; adjust the time range to see results"
@@ -24,6 +25,7 @@ const traceOutsideRangeWarn = "Query %s references a trace_id that exists betwee
 type builderQuery[T any] struct {
 	logger         *slog.Logger
 	telemetryStore telemetrystore.TelemetryStore
+	orgID          valuer.UUID
 	stmtBuilder    qbtypes.StatementBuilder[T]
 	spec           qbtypes.QueryBuilderQuery[T]
 	variables      map[string]qbtypes.VariableItem
@@ -45,6 +47,7 @@ type builderConfig struct {
 func newBuilderQuery[T any](
 	logger *slog.Logger,
 	telemetryStore telemetrystore.TelemetryStore,
+	orgID valuer.UUID,
 	stmtBuilder qbtypes.StatementBuilder[T],
 	spec qbtypes.QueryBuilderQuery[T],
 	tr qbtypes.TimeRange,
@@ -55,6 +58,7 @@ func newBuilderQuery[T any](
 	return &builderQuery[T]{
 		logger:         logger,
 		telemetryStore: telemetryStore,
+		orgID:          orgID,
 		stmtBuilder:    stmtBuilder,
 		spec:           spec,
 		variables:      variables,
@@ -214,7 +218,7 @@ func (q *builderQuery[T]) isWindowList() bool {
 
 // Statement renders the SQL without executing it, for the preview path.
 func (q *builderQuery[T]) Statement(ctx context.Context) (*qbtypes.Statement, error) {
-	return q.stmtBuilder.Build(ctx, q.fromMS, q.toMS, q.kind, q.spec, q.variables)
+	return q.stmtBuilder.Build(ctx, q.orgID, q.fromMS, q.toMS, q.kind, q.spec, q.variables)
 }
 
 func (q *builderQuery[T]) Execute(ctx context.Context) (*qbtypes.Result, error) {
@@ -238,7 +242,7 @@ func (q *builderQuery[T]) Execute(ctx context.Context) (*qbtypes.Result, error)
 		}
 	}
 
-	stmt, err := q.stmtBuilder.Build(ctx, fromMS, toMS, q.kind, q.spec, q.variables)
+	stmt, err := q.stmtBuilder.Build(ctx, q.orgID, fromMS, toMS, q.kind, q.spec, q.variables)
 	if err != nil {
 		return nil, err
 	}
@@ -491,7 +495,7 @@ func (q *builderQuery[T]) executeWindowList(ctx context.Context) (*qbtypes.Resul
 		q.spec.Offset = 0
 		q.spec.Limit = need
 
-		stmt, err := q.stmtBuilder.Build(ctx, r.fromNS/1e6, r.toNS/1e6, q.kind, q.spec, q.variables)
+		stmt, err := q.stmtBuilder.Build(ctx, q.orgID, r.fromNS/1e6, r.toNS/1e6, q.kind, q.spec, q.variables)
 		if err != nil {
 			return nil, err
 		}

pkg/querier/consume.go

@@ -21,6 +21,11 @@ import (
 
 var (
 	aggRe = regexp.MustCompile(`^__result_(\d+)$`)
+	// keyAliasRe matches the traces statement builder's positional column-alias prefix
+	// `__SELECT_KEY_<n>_` / `__GROUP_BY_KEY_<n>_`, which disambiguates select/group-by
+	// aliases from real table columns in the generated SQL. It is stripped here so the
+	// original field name surfaces as the label / column / raw-data key.
+	keyAliasRe = regexp.MustCompile(`^__(?:SELECT|GROUP_BY)_KEY_\d+_`)
 	// legacyReservedColumnTargetAliases identifies result value from a user
 	// written clickhouse query. The column alias indcate which value is
 	// to be considered as final result (or target).
@@ -29,6 +34,12 @@ var (
 	CodeFailUnmarshalJSONColumn = errors.MustNewCode("fail_unmarshal_json_column")
 )
 
+// stripKeyAlias removes the __SELECT_KEY_<n>_ / __GROUP_BY_KEY_<n>_ prefix from a result
+// column name, recovering the field name; unprefixed names are returned unchanged.
+func stripKeyAlias(name string) string {
+	return keyAliasRe.ReplaceAllString(name, "")
+}
+
 // consume reads every row and shapes it into the payload expected for the
 // given request type.
 //
@@ -126,7 +137,7 @@ func readAsTimeSeries(rows driver.Rows, queryWindow *qbtypes.TimeRange, step qbt
 		)
 
 		for idx, ptr := range slots {
-			name := colNames[idx]
+			name := stripKeyAlias(colNames[idx])
 
 			switch v := ptr.(type) {
 			case *time.Time:
@@ -296,6 +307,7 @@ func readAsScalar(rows driver.Rows, queryName string) (*qbtypes.ScalarData, erro
 
 	var aggIndex int64
 	for i, name := range colNames {
+		name = stripKeyAlias(name)
 		colType := qbtypes.ColumnTypeGroup
 		// Builder queries aliases aggregation columns as __result_N (always numeric) and wraps group-by keys with toString (always string);
 		// Raw ClickHouse queries may use any aliases.
@@ -406,7 +418,7 @@ func readAsRaw(rows driver.Rows, queryName string) (*qbtypes.RawData, error) {
 		}
 
 		for i, cellPtr := range scan {
-			name := colNames[i]
+			name := stripKeyAlias(colNames[i])
 
 			// de-reference the typed pointer to any
 			val := reflect.ValueOf(cellPtr).Elem().Interface()

pkg/querier/preview.go

@@ -78,7 +78,7 @@ func (q *querier) QueryRangePreview(
 			skip[name] = true
 		}
 	}
-	providers, buildErrs := q.buildPreviewProviders(req, dependencyQueries, missingMetricQuerySet, skip)
+	providers, buildErrs := q.buildPreviewProviders(orgID, req, dependencyQueries, missingMetricQuerySet, skip)
 
 	// Render each executing query's statement and collect the ClickHouse-bound
 	// analysis work to run concurrently.
@@ -192,6 +192,7 @@ func missingMetricNames(env qbtypes.QueryEnvelope) []string {
 }
 
 func (q *querier) buildPreviewProviders(
+	orgID valuer.UUID,
 	req *qbtypes.QueryRangeRequest,
 	dependencyQueries map[string]bool,
 	missingMetricQuerySet map[string]bool,
@@ -230,7 +231,7 @@ func (q *querier) buildPreviewProviders(
 			sub.CompositeQuery = qbtypes.CompositeQuery{Queries: []qbtypes.QueryEnvelope{query}}
 		}
 
-		built, _, bErr := q.buildQueries(&sub, deps, missingMetricQuerySet, event, promqlOptions{})
+		built, _, bErr := q.buildQueries(orgID, &sub, deps, missingMetricQuerySet, event)
 		if bErr != nil {
 			errs[name] = bErr
 			continue

pkg/querier/promql_query.go

@@ -8,19 +8,15 @@ import (
 	"regexp"
 	"sort"
 	"strings"
-	"sync"
 	"text/template"
 	"time"
 
-	"github.com/ClickHouse/clickhouse-go/v2"
-
 	"github.com/prometheus/prometheus/model/labels"
 	"github.com/prometheus/prometheus/promql"
 	"github.com/prometheus/prometheus/promql/parser"
 
 	"github.com/SigNoz/signoz/pkg/errors"
 	"github.com/SigNoz/signoz/pkg/prometheus"
-	"github.com/SigNoz/signoz/pkg/prometheus/clickhouseprometheusv2"
 	"github.com/SigNoz/signoz/pkg/querybuilder"
 	"github.com/SigNoz/signoz/pkg/types/ctxtypes"
 	"github.com/SigNoz/signoz/pkg/types/instrumentationtypes"
@@ -43,13 +39,6 @@ var quotedMetricOutsideBracesPattern = regexp.MustCompile(`"([^"]+)"\s*\{`)
 // tryEnhancePromQLExecError attempts to convert a PromQL execution error into
 // a properly typed error. Returns nil if the error is not a recognized execution error.
 func tryEnhancePromQLExecError(execErr error) error {
-	// A storage may fail a query with an already-typed error (e.g. the
-	// clickhousev2 series/sample budgets); surface it as-is instead of
-	// flattening it into an internal error.
-	if typed := typedStorageError(execErr); typed != nil {
-		return typed
-	}
-
 	var eqc promql.ErrQueryCanceled
 	var eqt promql.ErrQueryTimeout
 	var es promql.ErrStorage
@@ -69,30 +58,6 @@ func tryEnhancePromQLExecError(execErr error) error {
 	}
 }
 
-// typedStorageError walks an engine execution error chain looking for a
-// SigNoz-typed invalid-input error raised by the storage layer (the budget
-// refusals). Every wrapper level is stepped through by hand: Ast is a bare
-// type cast, not an unwrap — it misses a typed error behind the engine's
-// "expanding series: %w" — and promql.ErrStorage has no Unwrap method at
-// all, so a plain unwrap loop would stop at it.
-func typedStorageError(execErr error) error {
-	for e := execErr; e != nil; {
-		if errors.Ast(e, errors.TypeInvalidInput) {
-			return e
-		}
-		if es, ok := e.(promql.ErrStorage); ok {
-			e = es.Err
-			continue
-		}
-		u, ok := e.(interface{ Unwrap() error })
-		if !ok {
-			return nil
-		}
-		e = u.Unwrap()
-	}
-	return nil
-}
-
 // enhancePromQLError adds helpful context to PromQL parse errors,
 // particularly for UTF-8 syntax migration issues where metric and label
 // names containing dots need to be quoted.
@@ -133,24 +98,6 @@ type promqlQuery struct {
 	tr          qbv5.TimeRange
 	requestType qbv5.RequestType
 	vars        map[string]qbv5.VariableItem
-	opts        promqlOptions
-}
-
-// promqlOptions is how a PromQL query relates to the clickhousev2 provider
-// (see querier.promqlOptions for where the fields come from and why they are
-// flag-gated). Both providers are nil for a plain request, so a plain
-// request costs nothing extra.
-type promqlOptions struct {
-	// shadow, when set, runs the query on this provider after serving and
-	// logs any result difference; the response is never affected.
-	shadow *clickhouseprometheusv2.Provider
-	// shadowSlots is the querier-wide admission for shadow runs, shared by
-	// every query so the bound holds per process.
-	shadowSlots chan struct{}
-	// serve, when set, serves the response from this provider instead of the
-	// default path. Comparison callers fetch the default and the pinned
-	// result as two API calls and diff them.
-	serve *clickhouseprometheusv2.Provider
 }
 
 var _ qbv5.Query = (*promqlQuery)(nil)
@@ -163,7 +110,6 @@ func newPromqlQuery(
 	tr qbv5.TimeRange,
 	requestType qbv5.RequestType,
 	variables map[string]qbv5.VariableItem,
-	opts promqlOptions,
 ) *promqlQuery {
 	return &promqlQuery{
 		logger:      logger,
@@ -173,20 +119,10 @@ func newPromqlQuery(
 		tr:          tr,
 		requestType: requestType,
 		vars:        variables,
-		opts:        opts,
 	}
 }
 
 func (q *promqlQuery) Fingerprint() string {
-	// A pinned request must not share cache entries with default serving: a
-	// cached default result would satisfy the pin without running the pinned
-	// provider, and a pinned result would poison normal serving. No
-	// fingerprint means no caching at all — the pin exists to observe a
-	// provider, so a cache in front of it defeats the point.
-	if q.opts.serve != nil {
-		return ""
-	}
-
 	query, err := q.renderVars(q.query.Query, q.vars, q.tr.From, q.tr.To)
 	if err != nil {
 		q.logger.ErrorContext(context.TODO(), "failed render template variables", slog.String("query", q.query.Query))
@@ -312,16 +248,7 @@ func (q *promqlQuery) PreviewStatements(ctx context.Context) ([]prometheus.Captu
 	start := int64(querybuilder.ToNanoSecs(q.tr.From))
 	end := int64(querybuilder.ToNanoSecs(q.tr.To))
 
-	// Attach the same query traits as Execute so the captured statements
-	// match what the live path would run.
-	if expr, parseErr := q.parser.ParseExpr(rendered); parseErr == nil {
-		ctx = prometheus.NewContextWithQueryTraits(ctx, prometheus.DetectQueryTraits(expr))
-	}
-
 	capStorage, recorder := storer.CapturingStorage()
-	if capStorage == nil {
-		return nil, nil
-	}
 	qry, err := q.promEngine.Engine().NewRangeQuery(
 		ctx,
 		capStorage,
@@ -365,58 +292,6 @@ func (q *promqlQuery) Execute(ctx context.Context) (*qbv5.Result, error) {
 		return nil, err
 	}
 
-	// Attach query traits so the storage can prove step-aligned optimizations
-	// safe (see prometheus.QueryTraits). A parse failure surfaces below via
-	// the engine with the enhanced error message.
-	if expr, parseErr := q.parser.ParseExpr(query); parseErr == nil {
-		ctx = prometheus.NewContextWithQueryTraits(ctx, prometheus.DetectQueryTraits(expr))
-	}
-
-	// Accumulate ClickHouse-side scan stats across every storage query this
-	// evaluation issues (engine selectors or the compiled executor): progress
-	// options propagate to each ClickHouse query through the context.
-	var statsMu sync.Mutex
-	var rowsScanned, bytesScanned uint64
-	ctx = clickhouse.Context(ctx, clickhouse.WithProgress(func(p *clickhouse.Progress) {
-		statsMu.Lock()
-		rowsScanned += p.Rows
-		bytesScanned += p.Bytes
-		statsMu.Unlock()
-	}))
-
-	began := time.Now()
-
-	// A pinned provider serves directly from it: comparison callers fetch
-	// the default result and the pinned result as two API calls and diff
-	// them.
-	if q.opts.serve != nil {
-		matrix, err := q.serveFromProvider(ctx, query, start, end)
-		if err != nil {
-			if enhanced := tryEnhancePromQLExecError(err); enhanced != nil {
-				return nil, enhanced
-			}
-			return nil, err
-		}
-		return q.toResult(matrix, nil, began, &statsMu, &rowsScanned, &bytesScanned), nil
-	}
-
-	// When the serving provider itself is clickhousev2
-	// (prometheus::provider: clickhousev2), serve the way the provider is
-	// designed to serve: transpiled when the shape allows. Without this the
-	// override would silently run the engine path only.
-	if prov, ok := q.promEngine.(*clickhouseprometheusv2.Provider); ok {
-		matrix, served, err := prov.TryExecuteRange(ctx, query, time.Unix(0, start), time.Unix(0, end), q.query.Step.Duration)
-		if err != nil {
-			if enhanced := tryEnhancePromQLExecError(err); enhanced != nil {
-				return nil, enhanced
-			}
-			return nil, err
-		}
-		if served {
-			return q.toResult(matrix, nil, began, &statsMu, &rowsScanned, &bytesScanned), nil
-		}
-	}
-
 	qry, err := q.promEngine.Engine().NewRangeQuery(
 		ctx,
 		q.promEngine.Storage(),
@@ -452,34 +327,6 @@ func (q *promqlQuery) Execute(ctx context.Context) (*qbv5.Result, error) {
 		return nil, errors.WrapInternalf(promErr, errors.CodeInternal, "error getting matrix from promql query %q", query)
 	}
 
-	if q.opts.shadow != nil {
-		// Shadows detach from the request, so without admission a dashboard
-		// burst would stack unbounded ClickHouse work for up to the shadow
-		// timeout — the concurrency pattern behind the original outages.
-		// Non-blocking: at the cap the comparison is skipped, not queued;
-		// a sampled shadow stream is exactly as useful for rollout evidence.
-		select {
-		case q.opts.shadowSlots <- struct{}{}:
-			// The engine pools the result's sample slices on Close; the
-			// shadow comparison needs a stable copy of what was served.
-			served := copyMatrix(matrix)
-			servedIn := time.Since(began)
-			go func() {
-				defer func() { <-q.opts.shadowSlots }()
-				q.runShadowCompare(context.WithoutCancel(ctx), query, start, end, served, servedIn)
-			}()
-		default:
-			q.logger.DebugContext(ctx, "promql shadow skipped: at concurrency cap", slog.String("query", query))
-		}
-	}
-
-	warnings, _ := res.Warnings.AsStrings(query, 10, 0)
-	return q.toResult(matrix, warnings, began, &statsMu, &rowsScanned, &bytesScanned), nil
-}
-
-// toResult converts an evaluated matrix into the v5 result shape, attaching
-// the ClickHouse scan stats accumulated during evaluation.
-func (q *promqlQuery) toResult(matrix promql.Matrix, warnings []string, began time.Time, statsMu *sync.Mutex, rowsScanned, bytesScanned *uint64) *qbv5.Result {
 	excludeLabel := func(labelName string) bool {
 		if labelName == "__name__" {
 			return false
@@ -512,13 +359,7 @@ func (q *promqlQuery) toResult(matrix promql.Matrix, warnings []string, began ti
 		series = append(series, &s)
 	}
 
-	statsMu.Lock()
-	stats := qbv5.ExecStats{
-		RowsScanned:  *rowsScanned,
-		BytesScanned: *bytesScanned,
-		DurationMS:   uint64(time.Since(began).Milliseconds()),
-	}
-	statsMu.Unlock()
+	warnings, _ := res.Warnings.AsStrings(query, 10, 0)
 
 	return &qbv5.Result{
 		Type: q.requestType,
@@ -531,6 +372,6 @@ func (q *promqlQuery) toResult(matrix promql.Matrix, warnings []string, began ti
 			},
 		},
 		Warnings: warnings,
-		Stats:    stats,
-	}
+		// TODO: map promql stats?
+	}, nil
 }

pkg/querier/promql_query_test.go

@@ -7,9 +7,7 @@ import (
 
 	"github.com/SigNoz/signoz/pkg/errors"
 	"github.com/SigNoz/signoz/pkg/prometheus"
-	"github.com/SigNoz/signoz/pkg/prometheus/clickhouseprometheusv2"
 	qbv5 "github.com/SigNoz/signoz/pkg/types/querybuildertypes/querybuildertypesv5"
-	"github.com/prometheus/prometheus/promql"
 	"github.com/stretchr/testify/assert"
 )
 
@@ -442,35 +440,3 @@ func TestQuotedMetricOutsideBracesPattern(t *testing.T) {
 		})
 	}
 }
-
-// wrappedErr stands in for the engine's fmt-based "expanding series: %w"
-// wrapper: an ordinary error with an Unwrap chain that is not a SigNoz base
-// error itself.
-type wrappedErr struct{ inner error }
-
-func (w wrappedErr) Error() string { return "expanding series: " + w.inner.Error() }
-func (w wrappedErr) Unwrap() error { return w.inner }
-
-// A typed budget refusal must survive the engine's wrapping and reach the
-// API as invalid input; flattened to internal it becomes a 500 the user
-// cannot act on — the exact failure this error type exists to prevent.
-func TestTypedStorageError_SeesThroughEngineWrappers(t *testing.T) {
-	budget := errors.NewInvalidInputf(errors.CodeInvalidInput, "promql selector matched more than 500000 series")
-
-	assert.NotNil(t, typedStorageError(wrappedErr{inner: budget}), "typed error behind an Unwrap wrapper")
-	assert.NotNil(t, typedStorageError(promql.ErrStorage{Err: wrappedErr{inner: budget}}), "typed error behind ErrStorage then a wrapper")
-	assert.NotNil(t, typedStorageError(wrappedErr{inner: promql.ErrStorage{Err: budget}}), "typed error behind a wrapper then ErrStorage")
-	assert.Nil(t, typedStorageError(wrappedErr{inner: errors.NewInternalf(errors.CodeInternal, "boom")}), "internal errors stay internal")
-}
-
-// A pinned request must not share cache entries with default serving: a
-// cached default result would satisfy the pin without running the pinned
-// provider.
-func TestFingerprint_PinnedProviderBypassesCache(t *testing.T) {
-	q := &promqlQuery{
-		logger: slog.Default(),
-		query:  qbv5.PromQuery{Query: "up"},
-		opts:   promqlOptions{serve: &clickhouseprometheusv2.Provider{}},
-	}
-	assert.Empty(t, q.Fingerprint())
-}

pkg/querier/promql_shadow.go

@@ -1,195 +0,0 @@
-package querier
-
-import (
-	"context"
-	"fmt"
-	"log/slog"
-	"math"
-	"sort"
-	"time"
-
-	"github.com/ClickHouse/clickhouse-go/v2"
-	"github.com/SigNoz/signoz/pkg/prometheus"
-	"github.com/SigNoz/signoz/pkg/prometheus/clickhouseprometheusv2"
-	"github.com/prometheus/prometheus/model/labels"
-	"github.com/prometheus/prometheus/promql"
-)
-
-// shadowTimeout bounds a shadow evaluation; a shadow run must never outlive
-// the request by much or pile up.
-const shadowTimeout = 2 * time.Minute
-
-// runShadowCompare executes the query on the clickhousev2 provider exactly
-// as it would serve (transpiled when the shape allows, engine over the v2
-// querier otherwise), compares against the served result and logs the
-// outcome. Serving is never affected: this runs after the response, off the
-// request context, and only logs. The mismatch and failure logs are the
-// rollout evidence — serving cuts over to v2 only after they stay clean.
-func (q *promqlQuery) runShadowCompare(ctx context.Context, query string, startNs, endNs int64, served promql.Matrix, servedIn time.Duration) {
-	defer func() {
-		if r := recover(); r != nil {
-			q.logger.ErrorContext(ctx, "promql shadow comparison panicked", slog.Any("panic", r), slog.String("query", query))
-		}
-	}()
-
-	ctx, cancel := context.WithTimeout(ctx, shadowTimeout)
-	defer cancel()
-
-	// The request context carries the served response's scan-stats progress
-	// callback; without replacing it the shadow's ClickHouse progress would
-	// race into the served stats. The response itself was already sent.
-	ctx = clickhouse.Context(ctx, clickhouse.WithProgress(func(*clickhouse.Progress) {}))
-
-	if expr, parseErr := q.parser.ParseExpr(query); parseErr == nil {
-		ctx = prometheus.NewContextWithQueryTraits(ctx, prometheus.DetectQueryTraits(expr))
-	}
-
-	start, end := time.Unix(0, startNs), time.Unix(0, endNs)
-	began := time.Now()
-	shadow, transpiled, err := executeOnProvider(ctx, q.opts.shadow, query, start, end, q.query.Step.Duration)
-	shadowIn := time.Since(began)
-
-	logAttrs := []any{
-		slog.String("query", query),
-		slog.Int64("start_ms", startNs/int64(time.Millisecond)),
-		slog.Int64("end_ms", endNs/int64(time.Millisecond)),
-		slog.Duration("step", q.query.Step.Duration),
-		slog.Bool("transpiled", transpiled),
-		slog.Duration("served_in", servedIn),
-		slog.Duration("shadow_in", shadowIn),
-	}
-
-	if err != nil {
-		// A shadow failure would be a serving failure after rollout; surface
-		// it at the same level as a result mismatch.
-		q.logger.WarnContext(ctx, "promql shadow execution failed", append(logAttrs, slog.Any("error", err))...)
-		return
-	}
-
-	servedNorm := normalizeShadowMatrix(served)
-	shadowNorm := normalizeShadowMatrix(shadow)
-	if diff := diffShadowMatrices(servedNorm, shadowNorm); diff != "" {
-		q.logger.WarnContext(ctx, "promql shadow comparison mismatch", append(logAttrs,
-			slog.String("diff", diff),
-			slog.Int("served_series", len(servedNorm)),
-			slog.Int("shadow_series", len(shadowNorm)),
-		)...)
-		return
-	}
-	// Matches log the timings: served_in vs shadow_in across the fleet is
-	// the perf evidence for the cutover, gathered for free.
-	q.logger.DebugContext(ctx, "promql shadow comparison matched", logAttrs...)
-}
-
-// serveFromProvider evaluates the query the way the pinned provider would
-// serve it.
-func (q *promqlQuery) serveFromProvider(ctx context.Context, query string, startNs, endNs int64) (promql.Matrix, error) {
-	matrix, _, err := executeOnProvider(ctx, q.opts.serve, query, time.Unix(0, startNs), time.Unix(0, endNs), q.query.Step.Duration)
-	return matrix, err
-}
-
-// executeOnProvider evaluates the query the way the provider would serve it:
-// transpiled in ClickHouse when the shape allows, the engine over the
-// provider's storage otherwise. The returned matrix is an owned copy.
-func executeOnProvider(ctx context.Context, prov *clickhouseprometheusv2.Provider, query string, start, end time.Time, step time.Duration) (promql.Matrix, bool, error) {
-	matrix, ok, err := prov.TryExecuteRange(ctx, query, start, end, step)
-	if err != nil {
-		return nil, true, err
-	}
-	if ok {
-		return matrix, true, nil
-	}
-
-	qry, err := prov.Engine().NewRangeQuery(ctx, prov.Storage(), nil, query, start, end, step)
-	if err != nil {
-		return nil, false, err
-	}
-	defer qry.Close()
-
-	res := qry.Exec(ctx)
-	if res.Err != nil {
-		return nil, false, res.Err
-	}
-	matrix, err = res.Matrix()
-	if err != nil {
-		return nil, false, err
-	}
-	// Close returns the result's sample slices to the engine pool.
-	return copyMatrix(matrix), false, nil
-}
-
-func copyMatrix(matrix promql.Matrix) promql.Matrix {
-	out := make(promql.Matrix, 0, len(matrix))
-	for _, s := range matrix {
-		floats := make([]promql.FPoint, len(s.Floats))
-		copy(floats, s.Floats)
-		out = append(out, promql.Series{Metric: s.Metric.Copy(), Floats: floats})
-	}
-	return out
-}
-
-// normalizeShadowMatrix strips the labels the two providers legitimately
-// disagree on — v1 injects a synthetic fingerprint label and leaks
-// empty-valued labels from the stored attribute JSON, both removed from API
-// responses anyway — and sorts by label set.
-func normalizeShadowMatrix(matrix promql.Matrix) promql.Matrix {
-	out := make(promql.Matrix, 0, len(matrix))
-	for _, s := range matrix {
-		builder := labels.NewBuilder(s.Metric)
-		builder.Del(prometheus.FingerprintAsPromLabelName)
-		s.Metric.Range(func(l labels.Label) {
-			if l.Value == "" {
-				builder.Del(l.Name)
-			}
-		})
-		out = append(out, promql.Series{Metric: builder.Labels(), Floats: s.Floats})
-	}
-	sort.Slice(out, func(i, j int) bool { return labels.Compare(out[i].Metric, out[j].Metric) < 0 })
-	return out
-}
-
-// diffShadowMatrices returns a description of the first difference, or "".
-// Values compare with relative tolerance: spatial aggregations accumulate
-// floats in storage order, which differs between the providers in the last
-// ULP.
-func diffShadowMatrices(served, shadow promql.Matrix) string {
-	const relTol = 1e-9
-	if len(served) != len(shadow) {
-		return fmt.Sprintf("series count: served=%d shadow=%d", len(served), len(shadow))
-	}
-	for i := range served {
-		if labels.Compare(served[i].Metric, shadow[i].Metric) != 0 {
-			return fmt.Sprintf("series %d labels: served=%s shadow=%s", i, served[i].Metric, shadow[i].Metric)
-		}
-		if len(served[i].Floats) != len(shadow[i].Floats) {
-			return fmt.Sprintf("series %s points: served=%d shadow=%d", served[i].Metric, len(served[i].Floats), len(shadow[i].Floats))
-		}
-		for j := range served[i].Floats {
-			a, b := served[i].Floats[j], shadow[i].Floats[j]
-			if a.T != b.T {
-				return fmt.Sprintf("series %s point %d ts: served=%d shadow=%d", served[i].Metric, j, a.T, b.T)
-			}
-			// NaN and infinities first: NaN != NaN and Inf-Inf arithmetic
-			// would otherwise make one-sided NaN and Inf-vs-finite compare
-			// as equal (NaN > x and Inf > Inf are both false).
-			if math.IsNaN(a.F) || math.IsNaN(b.F) {
-				if math.IsNaN(a.F) != math.IsNaN(b.F) {
-					return fmt.Sprintf("series %s @%d value: served=%v shadow=%v", served[i].Metric, a.T, a.F, b.F)
-				}
-				continue
-			}
-			if math.IsInf(a.F, 0) || math.IsInf(b.F, 0) {
-				if a.F != b.F {
-					return fmt.Sprintf("series %s @%d value: served=%v shadow=%v", served[i].Metric, a.T, a.F, b.F)
-				}
-				continue
-			}
-			diff := math.Abs(a.F - b.F)
-			scale := math.Max(math.Abs(a.F), math.Abs(b.F))
-			if diff > relTol*math.Max(scale, 1e-300) && diff > 1e-12 {
-				return fmt.Sprintf("series %s @%d value: served=%v shadow=%v", served[i].Metric, a.T, a.F, b.F)
-			}
-		}
-	}
-	return ""
-}

pkg/querier/promql_shadow_test.go

@@ -1,66 +0,0 @@
-package querier
-
-import (
-	"math"
-	"testing"
-
-	"github.com/prometheus/prometheus/model/labels"
-	"github.com/prometheus/prometheus/promql"
-	"github.com/stretchr/testify/assert"
-)
-
-func TestNormalizeShadowMatrix(t *testing.T) {
-	matrix := promql.Matrix{
-		{
-			Metric: labels.FromStrings("__name__", "up", "fingerprint", "42", "empty", "", "job", "api"),
-			Floats: []promql.FPoint{{T: 1000, F: 1}},
-		},
-		{
-			Metric: labels.FromStrings("a", "1"),
-			Floats: []promql.FPoint{{T: 1000, F: 2}},
-		},
-	}
-	norm := normalizeShadowMatrix(matrix)
-	// sorted by labels; fingerprint and empty-valued labels stripped
-	assert.Equal(t, labels.FromStrings("__name__", "up", "job", "api"), norm[0].Metric)
-	assert.Equal(t, labels.FromStrings("a", "1"), norm[1].Metric)
-}
-
-func TestDiffShadowMatrices(t *testing.T) {
-	series := func(v float64) promql.Matrix {
-		return promql.Matrix{{Metric: labels.FromStrings("a", "1"), Floats: []promql.FPoint{{T: 1000, F: v}}}}
-	}
-
-	assert.Empty(t, diffShadowMatrices(series(1.5), series(1.5)))
-	// last-ULP differences from storage-order float accumulation are expected
-	assert.Empty(t, diffShadowMatrices(series(0.08888888888888889), series(0.08888888888888888)))
-	assert.Empty(t, diffShadowMatrices(series(math.NaN()), series(math.NaN())))
-
-	assert.Contains(t, diffShadowMatrices(series(1.5), series(1.6)), "value")
-	assert.Contains(t, diffShadowMatrices(series(1.5), promql.Matrix{}), "series count")
-	assert.Contains(t, diffShadowMatrices(
-		series(1.5),
-		promql.Matrix{{Metric: labels.FromStrings("a", "2"), Floats: []promql.FPoint{{T: 1000, F: 1.5}}}},
-	), "labels")
-	assert.Contains(t, diffShadowMatrices(
-		series(1.5),
-		promql.Matrix{{Metric: labels.FromStrings("a", "1"), Floats: []promql.FPoint{{T: 2000, F: 1.5}}}},
-	), "ts")
-}
-
-// One-sided NaN makes every float comparison false, and Inf-Inf arithmetic
-// yields Inf > Inf == false; without explicit handling both divergences log
-// as matched — a shadow comparator that cannot see them would green-light a
-// broken rollout.
-func TestDiffShadowMatrices_SpecialFloats(t *testing.T) {
-	point := func(v float64) promql.Matrix {
-		return promql.Matrix{{Metric: labels.FromStrings("a", "1"), Floats: []promql.FPoint{{T: 1000, F: v}}}}
-	}
-
-	assert.NotEmpty(t, diffShadowMatrices(point(math.NaN()), point(1.5)), "one-sided NaN must diff")
-	assert.NotEmpty(t, diffShadowMatrices(point(1.5), point(math.NaN())), "one-sided NaN must diff either way")
-	assert.NotEmpty(t, diffShadowMatrices(point(math.Inf(1)), point(1.5)), "Inf vs finite must diff")
-	assert.NotEmpty(t, diffShadowMatrices(point(math.Inf(1)), point(math.Inf(-1))), "opposite infinities must diff")
-	assert.Empty(t, diffShadowMatrices(point(math.Inf(1)), point(math.Inf(1))), "equal infinities match")
-	assert.Empty(t, diffShadowMatrices(point(math.NaN()), point(math.NaN())), "both NaN match")
-}

pkg/querier/querier.go

@@ -19,12 +19,10 @@ import (
 	"github.com/SigNoz/signoz/pkg/factory"
 	"github.com/SigNoz/signoz/pkg/flagger"
 	"github.com/SigNoz/signoz/pkg/prometheus"
-	"github.com/SigNoz/signoz/pkg/prometheus/clickhouseprometheusv2"
 	"github.com/SigNoz/signoz/pkg/query-service/utils"
 	"github.com/SigNoz/signoz/pkg/querybuilder"
 	"github.com/SigNoz/signoz/pkg/telemetrystore"
 	"github.com/SigNoz/signoz/pkg/types/ctxtypes"
-	"github.com/SigNoz/signoz/pkg/types/featuretypes"
 	"github.com/SigNoz/signoz/pkg/types/instrumentationtypes"
 	"github.com/SigNoz/signoz/pkg/types/metrictypes"
 	"github.com/SigNoz/signoz/pkg/types/telemetrytypes"
@@ -38,20 +36,13 @@ var (
 )
 
 type querier struct {
-	logger         *slog.Logger
-	fl             flagger.Flagger
-	telemetryStore telemetrystore.TelemetryStore
-	metadataStore  telemetrytypes.MetadataStore
-	promEngine     prometheus.Prometheus
-	// promV2 is the clickhousev2 prometheus provider, wired only when the
-	// serving provider is the default one (nil otherwise). It reads the same
-	// ClickHouse data through a different implementation; PromQL queries
-	// shadow-compare against it behind the use_prometheus_clickhouse_v2 flag
-	// and can be pinned to it for a response (see promqlOptions). It never
-	// serves by default — that cutover happens only after the shadow logs
-	// stay clean.
-	promV2                   *clickhouseprometheusv2.Provider
+	logger                   *slog.Logger
+	fl                       flagger.Flagger
+	telemetryStore           telemetrystore.TelemetryStore
+	metadataStore            telemetrytypes.MetadataStore
+	promEngine               prometheus.Prometheus
 	traceStmtBuilder         qbtypes.StatementBuilder[qbtypes.TraceAggregation]
+	aiTraceStmtBuilder       qbtypes.StatementBuilder[qbtypes.TraceAggregation]
 	logStmtBuilder           qbtypes.StatementBuilder[qbtypes.LogAggregation]
 	auditStmtBuilder         qbtypes.StatementBuilder[qbtypes.LogAggregation]
 	metricStmtBuilder        qbtypes.StatementBuilder[qbtypes.MetricAggregation]
@@ -61,16 +52,8 @@ type querier struct {
 	liveDataRefresh          time.Duration
 	builderConfig            builderConfig
 	maxConcurrentQueries     int
-	// shadowSlots bounds concurrent shadow comparisons per process; shadows
-	// detach from their requests, so nothing else limits how many pile up.
-	shadowSlots chan struct{}
 }
 
-// maxConcurrentShadows is deliberately small: a shadow is a full extra
-// ClickHouse evaluation, and a sampled stream of comparisons is exactly as
-// useful for rollout evidence as an exhaustive one under load.
-const maxConcurrentShadows = 8
-
 var _ Querier = (*querier)(nil)
 
 func New(
@@ -78,8 +61,8 @@ func New(
 	telemetryStore telemetrystore.TelemetryStore,
 	metadataStore telemetrytypes.MetadataStore,
 	promEngine prometheus.Prometheus,
-	promV2 *clickhouseprometheusv2.Provider,
 	traceStmtBuilder qbtypes.StatementBuilder[qbtypes.TraceAggregation],
+	aiTraceStmtBuilder qbtypes.StatementBuilder[qbtypes.TraceAggregation],
 	logStmtBuilder qbtypes.StatementBuilder[qbtypes.LogAggregation],
 	auditStmtBuilder qbtypes.StatementBuilder[qbtypes.LogAggregation],
 	metricStmtBuilder qbtypes.StatementBuilder[qbtypes.MetricAggregation],
@@ -100,8 +83,8 @@ func New(
 		telemetryStore:           telemetryStore,
 		metadataStore:            metadataStore,
 		promEngine:               promEngine,
-		promV2:                   promV2,
 		traceStmtBuilder:         traceStmtBuilder,
+		aiTraceStmtBuilder:       aiTraceStmtBuilder,
 		logStmtBuilder:           logStmtBuilder,
 		auditStmtBuilder:         auditStmtBuilder,
 		metricStmtBuilder:        metricStmtBuilder,
@@ -113,7 +96,6 @@ func New(
 			logTraceIDWindowPaddingMS: uint64(logTraceIDWindowPadding.Milliseconds()),
 		},
 		maxConcurrentQueries: maxConcurrentQueries,
-		shadowSlots:          make(chan struct{}, maxConcurrentShadows),
 	}
 }
 
@@ -153,11 +135,7 @@ func (q *querier) QueryRange(ctx context.Context, orgID valuer.UUID, req *qbtype
 		missingMetricQuerySet[name] = true
 	}
 
-	promqlOpts, err := q.promqlOptions(ctx, orgID, req)
-	if err != nil {
-		return nil, err
-	}
-	queries, steps, err := q.buildQueries(req, dependencyQueries, missingMetricQuerySet, event, promqlOpts)
+	queries, steps, err := q.buildQueries(orgID, req, dependencyQueries, missingMetricQuerySet, event)
 	if err != nil {
 		return nil, err
 	}
@@ -200,40 +178,12 @@ func (q *querier) QueryRange(ctx context.Context, orgID valuer.UUID, req *qbtype
 	return qbResp, qbErr
 }
 
-// promqlOptions derives the PromQL execution options for a request. With the
-// org's use_prometheus_clickhouse_v2 flag on, queries are shadow-compared
-// against the clickhousev2 provider (serving unaffected, diffs logged; see
-// promql_shadow.go). The X-SigNoz-PromQL-Provider header may instead pin the
-// response to that provider — integration tests and support fetch both
-// results for comparison — so it is deliberately flag-gated too: without the
-// gate the header would be an unaudited switch onto a provider still under
-// validation.
-func (q *querier) promqlOptions(ctx context.Context, orgID valuer.UUID, req *qbtypes.QueryRangeRequest) (promqlOptions, error) {
-	enabled := q.fl.BooleanOrEmpty(ctx, flagger.FeatureUsePrometheusClickhouseV2, featuretypes.NewFlaggerEvaluationContext(orgID))
-	if req.PromQLProvider == "" {
-		if enabled && q.promV2 != nil {
-			return promqlOptions{shadow: q.promV2, shadowSlots: q.shadowSlots}, nil
-		}
-		return promqlOptions{}, nil
-	}
-	if req.PromQLProvider != prometheus.ProviderClickhouseV2 {
-		return promqlOptions{}, errors.NewInvalidInputf(errors.CodeInvalidInput, "unknown promql provider %q", req.PromQLProvider)
-	}
-	if !enabled {
-		return promqlOptions{}, errors.NewInvalidInputf(errors.CodeInvalidInput, "promql provider %q requires the use_prometheus_clickhouse_v2 flag", req.PromQLProvider)
-	}
-	if q.promV2 == nil {
-		return promqlOptions{}, errors.NewInvalidInputf(errors.CodeInvalidInput, "promql provider %q is not available", req.PromQLProvider)
-	}
-	return promqlOptions{serve: q.promV2}, nil
-}
-
 func (q *querier) buildQueries(
+	orgID valuer.UUID,
 	req *qbtypes.QueryRangeRequest,
 	dependencyQueries map[string]bool,
 	missingMetricQuerySet map[string]bool,
 	event *qbtypes.QBEvent,
-	promqlOpts promqlOptions,
 ) (map[string]qbtypes.Query, map[string]qbtypes.Step, error) {
 
 	tmplVars := req.Variables
@@ -258,7 +208,7 @@ func (q *querier) buildQueries(
 			if !ok {
 				return nil, nil, errors.NewInvalidInputf(errors.CodeInvalidInput, "invalid promql query spec %T", query.Spec)
 			}
-			promqlQuery := newPromqlQuery(q.logger, q.promEngine, promQuery, qbtypes.TimeRange{From: req.Start, To: req.End}, req.RequestType, tmplVars, promqlOpts)
+			promqlQuery := newPromqlQuery(q.logger, q.promEngine, promQuery, qbtypes.TimeRange{From: req.Start, To: req.End}, req.RequestType, tmplVars)
 			queries[promQuery.Name] = promqlQuery
 			steps[promQuery.Name] = promQuery.Step
 		case qbtypes.QueryTypeClickHouseSQL:
@@ -275,6 +225,7 @@ func (q *querier) buildQueries(
 			}
 			toq := &traceOperatorQuery{
 				telemetryStore: q.telemetryStore,
+				orgID:          orgID,
 				stmtBuilder:    q.traceOperatorStmtBuilder,
 				spec:           traceOpQuery,
 				compositeQuery: &req.CompositeQuery,
@@ -289,7 +240,12 @@ func (q *querier) buildQueries(
 			case qbtypes.QueryBuilderQuery[qbtypes.TraceAggregation]:
 				spec.ShiftBy = extractShiftFromBuilderQuery(spec)
 				timeRange := adjustTimeRangeForShift(spec, qbtypes.TimeRange{From: req.Start, To: req.End}, req.RequestType)
-				bq := newBuilderQuery(q.logger, q.telemetryStore, q.traceStmtBuilder, spec, timeRange, req.RequestType, tmplVars, builderConfig{})
+				stmtBuilder := q.traceStmtBuilder
+				if spec.Source == telemetrytypes.SourceAI {
+					event.Source = telemetrytypes.SourceAI.StringValue()
+					stmtBuilder = q.aiTraceStmtBuilder
+				}
+				bq := newBuilderQuery(q.logger, q.telemetryStore, orgID, stmtBuilder, spec, timeRange, req.RequestType, tmplVars, builderConfig{})
 				queries[spec.Name] = bq
 				steps[spec.Name] = spec.StepInterval
 			case qbtypes.QueryBuilderQuery[qbtypes.LogAggregation]:
@@ -299,7 +255,7 @@ func (q *querier) buildQueries(
 				if spec.Source == telemetrytypes.SourceAudit {
 					stmtBuilder = q.auditStmtBuilder
 				}
-				bq := newBuilderQuery(q.logger, q.telemetryStore, stmtBuilder, spec, timeRange, req.RequestType, tmplVars, q.builderConfig)
+				bq := newBuilderQuery(q.logger, q.telemetryStore, orgID, stmtBuilder, spec, timeRange, req.RequestType, tmplVars, q.builderConfig)
 				queries[spec.Name] = bq
 				steps[spec.Name] = spec.StepInterval
 			case qbtypes.QueryBuilderQuery[qbtypes.MetricAggregation]:
@@ -316,9 +272,9 @@ func (q *querier) buildQueries(
 
 				if spec.Source == telemetrytypes.SourceMeter {
 					event.Source = telemetrytypes.SourceMeter.StringValue()
-					bq = newBuilderQuery(q.logger, q.telemetryStore, q.meterStmtBuilder, spec, timeRange, req.RequestType, tmplVars, builderConfig{})
+					bq = newBuilderQuery(q.logger, q.telemetryStore, orgID, q.meterStmtBuilder, spec, timeRange, req.RequestType, tmplVars, builderConfig{})
 				} else {
-					bq = newBuilderQuery(q.logger, q.telemetryStore, q.metricStmtBuilder, spec, timeRange, req.RequestType, tmplVars, builderConfig{})
+					bq = newBuilderQuery(q.logger, q.telemetryStore, orgID, q.metricStmtBuilder, spec, timeRange, req.RequestType, tmplVars, builderConfig{})
 				}
 
 				queries[spec.Name] = bq
@@ -589,7 +545,7 @@ func (q *querier) QueryRawStream(ctx context.Context, orgID valuer.UUID, req *qb
 			if spec.Source == telemetrytypes.SourceAudit {
 				liveTailStmtBuilder = q.auditStmtBuilder
 			}
-			bq := newBuilderQuery(q.logger, q.telemetryStore, liveTailStmtBuilder, spec, timeRange, req.RequestType, map[string]qbtypes.VariableItem{
+			bq := newBuilderQuery(q.logger, q.telemetryStore, orgID, liveTailStmtBuilder, spec, timeRange, req.RequestType, map[string]qbtypes.VariableItem{
 				"id": {
 					Value: updatedLogID,
 				},
@@ -900,7 +856,7 @@ func (q *querier) createRangedQuery(originalQuery qbtypes.Query, timeRange qbtyp
 	switch qt := originalQuery.(type) {
 	case *promqlQuery:
 		queryCopy := qt.query.Copy()
-		return newPromqlQuery(q.logger, qt.promEngine, queryCopy, timeRange, qt.requestType, qt.vars, qt.opts)
+		return newPromqlQuery(q.logger, q.promEngine, queryCopy, timeRange, qt.requestType, qt.vars)
 
 	case *chSQLQuery:
 		queryCopy := qt.query.Copy()
@@ -912,7 +868,11 @@ func (q *querier) createRangedQuery(originalQuery qbtypes.Query, timeRange qbtyp
 		specCopy := qt.spec.Copy()
 		specCopy.ShiftBy = extractShiftFromBuilderQuery(specCopy)
 		adjustedTimeRange := adjustTimeRangeForShift(specCopy, timeRange, qt.kind)
-		return newBuilderQuery(q.logger, q.telemetryStore, q.traceStmtBuilder, specCopy, adjustedTimeRange, qt.kind, qt.variables, builderConfig{})
+		shiftStmtBuilder := q.traceStmtBuilder
+		if qt.spec.Source == telemetrytypes.SourceAI {
+			shiftStmtBuilder = q.aiTraceStmtBuilder
+		}
+		return newBuilderQuery(q.logger, q.telemetryStore, qt.orgID, shiftStmtBuilder, specCopy, adjustedTimeRange, qt.kind, qt.variables, builderConfig{})
 
 	case *builderQuery[qbtypes.LogAggregation]:
 		specCopy := qt.spec.Copy()
@@ -922,20 +882,21 @@ func (q *querier) createRangedQuery(originalQuery qbtypes.Query, timeRange qbtyp
 		if qt.spec.Source == telemetrytypes.SourceAudit {
 			shiftStmtBuilder = q.auditStmtBuilder
 		}
-		return newBuilderQuery(q.logger, q.telemetryStore, shiftStmtBuilder, specCopy, adjustedTimeRange, qt.kind, qt.variables, q.builderConfig)
+		return newBuilderQuery(q.logger, q.telemetryStore, qt.orgID, shiftStmtBuilder, specCopy, adjustedTimeRange, qt.kind, qt.variables, q.builderConfig)
 
 	case *builderQuery[qbtypes.MetricAggregation]:
 		specCopy := qt.spec.Copy()
 		specCopy.ShiftBy = extractShiftFromBuilderQuery(specCopy)
 		adjustedTimeRange := adjustTimeRangeForShift(specCopy, timeRange, qt.kind)
 		if qt.spec.Source == telemetrytypes.SourceMeter {
-			return newBuilderQuery(q.logger, q.telemetryStore, q.meterStmtBuilder, specCopy, adjustedTimeRange, qt.kind, qt.variables, builderConfig{})
+			return newBuilderQuery(q.logger, q.telemetryStore, qt.orgID, q.meterStmtBuilder, specCopy, adjustedTimeRange, qt.kind, qt.variables, builderConfig{})
 		}
-		return newBuilderQuery(q.logger, q.telemetryStore, q.metricStmtBuilder, specCopy, adjustedTimeRange, qt.kind, qt.variables, builderConfig{})
+		return newBuilderQuery(q.logger, q.telemetryStore, qt.orgID, q.metricStmtBuilder, specCopy, adjustedTimeRange, qt.kind, qt.variables, builderConfig{})
 	case *traceOperatorQuery:
 		specCopy := qt.spec.Copy()
 		return &traceOperatorQuery{
 			telemetryStore: q.telemetryStore,
+			orgID:          qt.orgID,
 			stmtBuilder:    q.traceOperatorStmtBuilder,
 			spec:           specCopy,
 			fromMS:         uint64(timeRange.From),

pkg/querier/querier_test.go

@@ -30,7 +30,7 @@ func (m *queryMatcherAny) Match(string, string) error { return nil }
 // and returns a fixed query string so the mock ClickHouse can match it.
 type mockMetricStmtBuilder struct{}
 
-func (m *mockMetricStmtBuilder) Build(_ context.Context, _, _ uint64, _ qbtypes.RequestType, _ qbtypes.QueryBuilderQuery[qbtypes.MetricAggregation], _ map[string]qbtypes.VariableItem) (*qbtypes.Statement, error) {
+func (m *mockMetricStmtBuilder) Build(_ context.Context, _ valuer.UUID, _, _ uint64, _ qbtypes.RequestType, _ qbtypes.QueryBuilderQuery[qbtypes.MetricAggregation], _ map[string]qbtypes.VariableItem) (*qbtypes.Statement, error) {
 	return &qbtypes.Statement{
 		Query: "SELECT ts, value FROM signoz_metrics",
 		Args:  nil,
@@ -48,8 +48,8 @@ func TestQueryRange_MetricTypeMissing(t *testing.T) {
 		nil, // telemetryStore
 		metadataStore,
 		nil,                // prometheus
-		nil,                // promV2
 		nil,                // traceStmtBuilder
+		nil,                // aiTraceStmtBuilder
 		nil,                // logStmtBuilder
 		nil,                // auditStmtBuilder
 		nil,                // metricStmtBuilder
@@ -121,8 +121,8 @@ func TestQueryRange_MetricTypeFromStore(t *testing.T) {
 		telemetryStore,
 		metadataStore,
 		nil,                      // prometheus
-		nil,                      // promV2
 		nil,                      // traceStmtBuilder
+		nil,                      // aiTraceStmtBuilder
 		nil,                      // logStmtBuilder
 		nil,                      // auditStmtBuilder
 		&mockMetricStmtBuilder{}, // metricStmtBuilder

pkg/querier/signozquerier/provider.go

@@ -7,9 +7,9 @@ import (
 	"github.com/SigNoz/signoz/pkg/factory"
 	"github.com/SigNoz/signoz/pkg/flagger"
 	"github.com/SigNoz/signoz/pkg/prometheus"
-	"github.com/SigNoz/signoz/pkg/prometheus/clickhouseprometheusv2"
 	"github.com/SigNoz/signoz/pkg/querier"
 	"github.com/SigNoz/signoz/pkg/querybuilder"
+	"github.com/SigNoz/signoz/pkg/telemetryai"
 	"github.com/SigNoz/signoz/pkg/telemetryaudit"
 	"github.com/SigNoz/signoz/pkg/telemetrylogs"
 	"github.com/SigNoz/signoz/pkg/telemetrymetadata"
@@ -23,7 +23,6 @@ import (
 func NewFactory(
 	telemetryStore telemetrystore.TelemetryStore,
 	prometheus prometheus.Prometheus,
-	promV2 *clickhouseprometheusv2.Provider,
 	cache cache.Cache,
 	flagger flagger.Flagger,
 ) factory.ProviderFactory[querier.Querier, querier.Config] {
@@ -34,7 +33,7 @@ func NewFactory(
 			settings factory.ProviderSettings,
 			cfg querier.Config,
 		) (querier.Querier, error) {
-			return newProvider(ctx, settings, cfg, telemetryStore, prometheus, promV2, cache, flagger)
+			return newProvider(ctx, settings, cfg, telemetryStore, prometheus, cache, flagger)
 		},
 	)
 }
@@ -45,7 +44,6 @@ func newProvider(
 	cfg querier.Config,
 	telemetryStore telemetrystore.TelemetryStore,
 	prometheus prometheus.Prometheus,
-	promV2 *clickhouseprometheusv2.Provider,
 	cache cache.Cache,
 	flagger flagger.Flagger,
 ) (querier.Querier, error) {
@@ -95,6 +93,17 @@ func newProvider(
 		cfg.SkipResourceFingerprint.Threshold,
 	)
 
+	// AI trace statement builder (source=ai). The gen_ai gate/column keys are
+	// surfaced by the metadata store itself (enrichWithGenAIKeys), so queries work
+	// before any gen_ai metadata is ingested — no per-builder decoration needed.
+	// The standard trace builder doubles as the delegate for the span-list path.
+	aiTraceStmtBuilder := telemetryai.NewAITraceStatementBuilder(
+		settings,
+		telemetryMetadataStore,
+		traceStmtBuilder,
+		flagger,
+	)
+
 	// Create trace operator statement builder
 	traceOperatorStmtBuilder := telemetrytraces.NewTraceOperatorStatementBuilder(
 		settings,
@@ -187,8 +196,8 @@ func newProvider(
 		telemetryStore,
 		telemetryMetadataStore,
 		prometheus,
-		promV2,
 		traceStmtBuilder,
+		aiTraceStmtBuilder,
 		logStmtBuilder,
 		auditStmtBuilder,
 		metricStmtBuilder,

pkg/querier/trace_operator_query.go

@@ -10,10 +10,12 @@ import (
 	"github.com/SigNoz/signoz/pkg/types/instrumentationtypes"
 	qbtypes "github.com/SigNoz/signoz/pkg/types/querybuildertypes/querybuildertypesv5"
 	"github.com/SigNoz/signoz/pkg/types/telemetrytypes"
+	"github.com/SigNoz/signoz/pkg/valuer"
 )
 
 type traceOperatorQuery struct {
 	telemetryStore telemetrystore.TelemetryStore
+	orgID          valuer.UUID
 	stmtBuilder    qbtypes.TraceOperatorStatementBuilder
 	spec           qbtypes.QueryBuilderTraceOperator
 	compositeQuery *qbtypes.CompositeQuery
@@ -35,12 +37,13 @@ func (q *traceOperatorQuery) Window() (uint64, uint64) {
 
 // Statement renders the SQL without executing it, for the preview path.
 func (q *traceOperatorQuery) Statement(ctx context.Context) (*qbtypes.Statement, error) {
-	return q.stmtBuilder.Build(ctx, q.fromMS, q.toMS, q.kind, q.spec, q.compositeQuery)
+	return q.stmtBuilder.Build(ctx, q.orgID, q.fromMS, q.toMS, q.kind, q.spec, q.compositeQuery)
 }
 
 func (q *traceOperatorQuery) Execute(ctx context.Context) (*qbtypes.Result, error) {
 	stmt, err := q.stmtBuilder.Build(
 		ctx,
+		q.orgID,
 		q.fromMS,
 		q.toMS,
 		q.kind,

pkg/query-service/rules/managertestfactory.go

@@ -105,7 +105,7 @@ func NewTestManager(t *testing.T, testOpts *TestManagerOptions) *Manager {
 	}
 
 	// Create querier with test values
-	providerFactory := signozquerier.NewFactory(telemetryStore, prometheus, nil, cache, flagger)
+	providerFactory := signozquerier.NewFactory(telemetryStore, prometheus, cache, flagger)
 	mockQuerier, err := providerFactory.New(context.Background(), providerSettings, querier.Config{})
 	require.NoError(t, err)
 

pkg/query-service/rules/setups_test.go

@@ -47,8 +47,8 @@ func prepareQuerierForMetrics(t *testing.T, telemetryStore telemetrystore.Teleme
 		telemetryStore,
 		metadataStore,
 		nil, // prometheus
-		nil, // promV2
 		nil, // traceStmtBuilder
+		nil, // aiTraceStmtBuilder
 		nil, // logStmtBuilder
 		nil, // auditStmtBuilder
 		metricStmtBuilder,
@@ -103,8 +103,8 @@ func prepareQuerierForLogs(t *testing.T, telemetryStore telemetrystore.Telemetry
 		telemetryStore,
 		metadataStore,
 		nil,            // prometheus
-		nil,            // promV2
 		nil,            // traceStmtBuilder
+		nil,            // aiTraceStmtBuilder
 		logStmtBuilder, // logStmtBuilder
 		nil,            // auditStmtBuilder
 		nil,            // metricStmtBuilder
@@ -153,8 +153,8 @@ func prepareQuerierForTraces(t *testing.T, telemetryStore telemetrystore.Telemet
 		telemetryStore,
 		metadataStore,
 		nil,              // prometheus
-		nil,              // promV2
 		traceStmtBuilder, // traceStmtBuilder
+		nil,              // aiTraceStmtBuilder
 		nil,              // logStmtBuilder
 		nil,              // auditStmtBuilder
 		nil,              // metricStmtBuilder

pkg/querybuilder/agg_rewrite.go

@@ -53,6 +53,7 @@ func NewAggExprRewriter(
 // and the args if the parametric aggregation function is used.
 func (r *aggExprRewriter) Rewrite(
 	ctx context.Context,
+	orgID valuer.UUID,
 	startNs uint64,
 	endNs uint64,
 	expr string,
@@ -83,6 +84,7 @@ func (r *aggExprRewriter) Rewrite(
 
 	visitor := newExprVisitor(
 		ctx,
+		orgID,
 		startNs,
 		endNs,
 		r.logger,
@@ -107,6 +109,7 @@ func (r *aggExprRewriter) Rewrite(
 // RewriteMulti rewrites a slice of expressions.
 func (r *aggExprRewriter) RewriteMulti(
 	ctx context.Context,
+	orgID valuer.UUID,
 	startNs uint64,
 	endNs uint64,
 	exprs []string,
@@ -117,7 +120,7 @@ func (r *aggExprRewriter) RewriteMulti(
 	var errs []error
 	var chArgsList [][]any
 	for i, e := range exprs {
-		w, chArgs, err := r.Rewrite(ctx, startNs, endNs, e, rateInterval, keys)
+		w, chArgs, err := r.Rewrite(ctx, orgID, startNs, endNs, e, rateInterval, keys)
 		if err != nil {
 			errs = append(errs, err)
 			out[i] = e
@@ -135,6 +138,7 @@ func (r *aggExprRewriter) RewriteMulti(
 // exprVisitor walks FunctionExpr nodes and applies the mappers.
 type exprVisitor struct {
 	ctx     context.Context
+	orgID   valuer.UUID
 	startNs uint64
 	endNs   uint64
 	chparser.DefaultASTVisitor
@@ -152,6 +156,7 @@ type exprVisitor struct {
 
 func newExprVisitor(
 	ctx context.Context,
+	orgID valuer.UUID,
 	startNs uint64,
 	endNs uint64,
 	logger *slog.Logger,
@@ -164,6 +169,7 @@ func newExprVisitor(
 ) *exprVisitor {
 	return &exprVisitor{
 		ctx:              ctx,
+		orgID:            orgID,
 		startNs:          startNs,
 		endNs:            endNs,
 		logger:           logger,
@@ -206,7 +212,7 @@ func (v *exprVisitor) VisitFunctionExpr(fn *chparser.FunctionExpr) error {
 		dataType = telemetrytypes.FieldDataTypeFloat64
 	}
 
-	bodyJSONEnabled := v.flagger.BooleanOrEmpty(v.ctx, flagger.FeatureUseJSONBody, featuretypes.NewFlaggerEvaluationContext(valuer.UUID{}))
+	bodyJSONEnabled := v.flagger.BooleanOrEmpty(v.ctx, flagger.FeatureUseJSONBody, featuretypes.NewFlaggerEvaluationContext(v.orgID))
 
 	// Handle *If functions with predicate + values
 	if aggFunc.FuncCombinator {
@@ -216,6 +222,7 @@ func (v *exprVisitor) VisitFunctionExpr(fn *chparser.FunctionExpr) error {
 			origPred,
 			FilterExprVisitorOpts{
 				Context:          v.ctx,
+				OrgID:            v.orgID,
 				Logger:           v.logger,
 				FieldKeys:        v.fieldKeys,
 				FieldMapper:      v.fieldMapper,
@@ -247,7 +254,7 @@ func (v *exprVisitor) VisitFunctionExpr(fn *chparser.FunctionExpr) error {
 		for i := 0; i < len(args)-1; i++ {
 			origVal := args[i].String()
 			fieldKey := telemetrytypes.GetFieldKeyFromKeyText(origVal)
-			expr, exprArgs, err := CollisionHandledFinalExpr(v.ctx, v.startNs, v.endNs, &fieldKey, v.fieldMapper, v.conditionBuilder, v.fieldKeys, dataType, v.jsonKeyToKey, bodyJSONEnabled)
+			expr, exprArgs, err := CollisionHandledFinalExpr(v.ctx, v.orgID, v.startNs, v.endNs, &fieldKey, v.fieldMapper, v.conditionBuilder, v.fieldKeys, dataType, v.jsonKeyToKey, bodyJSONEnabled)
 			if err != nil {
 				return errors.WrapInvalidInputf(err, errors.CodeInvalidInput, "failed to get table field name for %q", origVal)
 			}
@@ -265,7 +272,7 @@ func (v *exprVisitor) VisitFunctionExpr(fn *chparser.FunctionExpr) error {
 		for i, arg := range args {
 			orig := arg.String()
 			fieldKey := telemetrytypes.GetFieldKeyFromKeyText(orig)
-			expr, exprArgs, err := CollisionHandledFinalExpr(v.ctx, v.startNs, v.endNs, &fieldKey, v.fieldMapper, v.conditionBuilder, v.fieldKeys, dataType, v.jsonKeyToKey, bodyJSONEnabled)
+			expr, exprArgs, err := CollisionHandledFinalExpr(v.ctx, v.orgID, v.startNs, v.endNs, &fieldKey, v.fieldMapper, v.conditionBuilder, v.fieldKeys, dataType, v.jsonKeyToKey, bodyJSONEnabled)
 			if err != nil {
 				return err
 			}

pkg/querybuilder/fallback_expr.go

@@ -13,12 +13,14 @@ import (
 	"github.com/SigNoz/signoz/pkg/errors"
 	qbtypes "github.com/SigNoz/signoz/pkg/types/querybuildertypes/querybuildertypesv5"
 	"github.com/SigNoz/signoz/pkg/types/telemetrytypes"
+	"github.com/SigNoz/signoz/pkg/valuer"
 	"github.com/huandu/go-sqlbuilder"
 	"golang.org/x/exp/maps"
 )
 
 func CollisionHandledFinalExpr(
 	ctx context.Context,
+	orgID valuer.UUID,
 	startNs uint64,
 	endNs uint64,
 	field *telemetrytypes.TelemetryFieldKey,
@@ -47,7 +49,7 @@ func CollisionHandledFinalExpr(
 
 	addCondition := func(key *telemetrytypes.TelemetryFieldKey) error {
 		sb := sqlbuilder.NewSelectBuilder()
-		conds, _, err := cb.ConditionFor(ctx, startNs, endNs, key, []*telemetrytypes.TelemetryFieldKey{key}, qbtypes.FilterOperatorExists, nil, sb)
+		conds, _, err := cb.ConditionFor(ctx, orgID, startNs, endNs, key, []*telemetrytypes.TelemetryFieldKey{key}, qbtypes.FilterOperatorExists, nil, sb)
 		if err != nil {
 			return err
 		}
@@ -76,23 +78,23 @@ func CollisionHandledFinalExpr(
 		}
 
 		if len(keysForField) == 0 {
-			// - the context is not provided
-			// - there are not keys for the field
-			// - it is not a static field
-			// - the next best thing to do is see if there is a typo
-			// and suggest a correction
+			// No metadata match: let the field mapper synthesize type-variant keys.
+			// keys were already checked above, so pass nil here.
+			keysForField = fm.CandidateKeys(ctx, orgID, field, nil, nil)
+		}
+		if len(keysForField) == 0 {
+			// The mapper can't synthesize (e.g. metrics): fall back to the typo suggestion.
 			wrappedErr := errors.WithSuggestiveAdditionalf(fieldForErr, errors.NewSuggestionsOnLevenshteinDistance(field.Name, errors.NounKeys, maps.Keys(keys)), "field `%s` not found", field.Name)
 			return "", nil, wrappedErr
-		} else {
-			for _, key := range keysForField {
-				err := addCondition(key)
-				if err != nil {
-					return "", nil, err
-				}
-				fieldExpression, _ = fm.FieldFor(ctx, startNs, endNs, key)
-				fieldExpression, _ = DataTypeCollisionHandledFieldName(key, dummyValue, fieldExpression, qbtypes.FilterOperatorUnknown)
-				stmts = append(stmts, fieldExpression)
+		}
+		for _, key := range keysForField {
+			err := addCondition(key)
+			if err != nil {
+				return "", nil, err
 			}
+			fieldExpression, _ = fm.FieldFor(ctx, startNs, endNs, key)
+			fieldExpression, _ = DataTypeCollisionHandledFieldName(key, dummyValue, fieldExpression, qbtypes.FilterOperatorUnknown)
+			stmts = append(stmts, fieldExpression)
 		}
 	} else {
 		err := addCondition(field)

pkg/querybuilder/filter_split.go

@@ -0,0 +1,154 @@
+package querybuilder
+
+import (
+	"strings"
+
+	"github.com/SigNoz/signoz/pkg/errors"
+	grammar "github.com/SigNoz/signoz/pkg/parser/filterquery/grammar"
+	"github.com/SigNoz/signoz/pkg/types/telemetrytypes"
+	"github.com/antlr4-go/antlr/v4"
+)
+
+// SplitFilterForAggregates partitions a single filter expression into a span-level
+// part (a WHERE over spans) and a trace-level part (a HAVING over per-trace
+// aggregates), splitting on the top-level AND.
+//
+// A key is trace-level when it carries the trace field context (`trace.completion_tokens`)
+// or, with no context, its bare name is in aggregateNames. Any other explicit context
+// (`span.`, `resource.`, …) is span-level. Trace-level and span-level keys may be
+// AND-combined (they run at different query stages) but not OR-combined; an OR that
+// mixes the two is an error.
+//
+// Syntax errors are ignored here — each part is re-parsed downstream (PrepareWhereClause
+// for the span part, the HAVING rewriter for the trace part), which surface them.
+func SplitFilterForAggregates(query string, aggregateNames map[string]struct{}) (spanExpr string, havingExpr string, err error) {
+	if strings.TrimSpace(query) == "" {
+		return "", "", nil
+	}
+
+	s := filterSplitter{query: []rune(query), aggregateNames: aggregateNames}
+	s.visit(parseFilterQuery(query))
+
+	if s.mixed {
+		return "", "", errors.NewInvalidInputf(errors.CodeInvalidInput,
+			"trace-level and span-level filters cannot be combined within an OR/NOT group; separate them with a top-level AND")
+	}
+	return strings.Join(s.span, " AND "), strings.Join(s.having, " AND "), nil
+}
+
+func parseFilterQuery(query string) antlr.Tree {
+	lexer := grammar.NewFilterQueryLexer(antlr.NewInputStream(query))
+	lexer.RemoveErrorListeners()
+	parser := grammar.NewFilterQueryParser(antlr.NewCommonTokenStream(lexer, 0))
+	parser.RemoveErrorListeners()
+	return parser.Query()
+}
+
+// filterSplitter walks the parse tree once, flattening the top-level AND chain and
+// routing each atom (a comparison, a NOT expression, or a whole multi-branch OR group)
+// to the span or having bucket by the class of the keys it references.
+type filterSplitter struct {
+	query          []rune
+	aggregateNames map[string]struct{}
+	span           []string
+	having         []string
+	mixed          bool
+}
+
+func (s *filterSplitter) visit(node antlr.Tree) {
+	switch n := node.(type) {
+	case *grammar.QueryContext:
+		if n.Expression() != nil {
+			s.visit(n.Expression())
+		}
+	case *grammar.ExpressionContext:
+		if n.OrExpression() != nil {
+			s.visit(n.OrExpression())
+		}
+	case *grammar.OrExpressionContext:
+		// a single branch is just an AND chain; multiple branches are a real OR, kept
+		// whole so a class-mixing OR can be rejected.
+		if ands := n.AllAndExpression(); len(ands) == 1 {
+			s.visit(ands[0])
+		} else {
+			s.route(n)
+		}
+	case *grammar.AndExpressionContext:
+		for _, u := range n.AllUnaryExpression() {
+			s.visit(u)
+		}
+	case *grammar.UnaryExpressionContext:
+		if n.NOT() != nil {
+			s.route(n)
+		} else if n.Primary() != nil {
+			s.visit(n.Primary())
+		}
+	case *grammar.PrimaryContext:
+		if n.OrExpression() != nil { // parenthesized sub-expression
+			s.visit(n.OrExpression())
+		} else {
+			s.route(n)
+		}
+	}
+}
+
+// route classifies an atom and appends its original source text to the right bucket.
+func (s *filterSplitter) route(atom antlr.ParserRuleContext) {
+	isTrace, isSpan := classifyKeys(atom, s.aggregateNames)
+	if isTrace && isSpan {
+		s.mixed = true
+		return
+	}
+	text := atomSourceText(s.query, atom)
+	// A multi-branch OR group's source slice excludes its enclosing parens (they belong
+	// to the parent Primary). Re-wrap it so rejoining a bucket with " AND " cannot invert
+	// OR/AND precedence, e.g. `a AND (b OR c)` must not flatten to `a AND b OR c`.
+	if or, ok := atom.(*grammar.OrExpressionContext); ok && len(or.AllAndExpression()) > 1 {
+		text = "(" + text + ")"
+	}
+	if isTrace {
+		s.having = append(s.having, text)
+	} else {
+		s.span = append(s.span, text)
+	}
+}
+
+// classifyKeys reports whether a subtree references trace-level and/or span-level keys.
+// A key is trace-level when it carries the trace field context or, with no context,
+// its name is a known aggregate; an unknown name under the trace context stays
+// trace-level so the aggregate validation rejects it with a targeted error. Any other
+// explicit context (`span.`, `resource.`, …) is span-level.
+func classifyKeys(node antlr.Tree, aggregateNames map[string]struct{}) (isTrace, isSpan bool) {
+	kc, ok := node.(*grammar.KeyContext)
+	if ok {
+		key := telemetrytypes.GetFieldKeyFromKeyText(kc.GetText())
+		switch key.FieldContext {
+		case telemetrytypes.FieldContextTrace:
+			isTrace = true
+		case telemetrytypes.FieldContextUnspecified:
+			_, isTrace = aggregateNames[key.Name]
+			isSpan = !isTrace
+		default:
+			isSpan = true
+		}
+		return
+	}
+	for i := 0; i < node.GetChildCount(); i++ {
+		t, s := classifyKeys(node.GetChild(i), aggregateNames)
+		isTrace = isTrace || t
+		isSpan = isSpan || s
+	}
+	return
+}
+
+// atomSourceText returns the original source substring for an atom, preserving
+// whitespace. The token stream drops skipped whitespace, which would glue word
+// operators (OR/AND/NOT) to their operands, so slice the input by token offsets.
+// ANTLR offsets are rune indices (InputStream holds []rune), hence the rune slice.
+func atomSourceText(query []rune, atom antlr.ParserRuleContext) string {
+	start, stop := atom.GetStart(), atom.GetStop()
+	if start == nil || stop == nil || start.GetStart() < 0 || stop.GetStop() >= len(query) || stop.GetStop() < start.GetStart() {
+		return atom.GetText()
+	}
+	return string(query[start.GetStart() : stop.GetStop()+1])
+}

pkg/querybuilder/filter_split_test.go

@@ -0,0 +1,167 @@
+package querybuilder
+
+import (
+	"testing"
+
+	"github.com/stretchr/testify/require"
+)
+
+func TestSplitFilterForAggregates(t *testing.T) {
+	agg := map[string]struct{}{"completion_tokens": {}, "span_count": {}, "prompt_tokens": {}}
+
+	type tc struct {
+		name    string
+		query   string
+		span    string // expected span-level (WHERE) part; "" => empty
+		having  string // expected trace-level (HAVING) part; "" => empty
+		wantErr bool
+	}
+
+	cases := []tc{
+		// --- empty input ---------------------------------------------------------
+		{
+			name: "empty",
+		},
+		{
+			name:  "whitespace only",
+			query: "   ",
+		},
+
+		// --- single class --------------------------------------------------------
+		{
+			name:  "span only",
+			query: "service.name = 'x'",
+			span:  "service.name = 'x'",
+		},
+		{
+			name:   "agg only bare",
+			query:  "completion_tokens > 1000",
+			having: "completion_tokens > 1000",
+		},
+		{
+			// the user-facing `trace.` prefix marks a trace-level aggregate.
+			name:   "agg only trace prefix",
+			query:  "trace.completion_tokens > 1000",
+			having: "trace.completion_tokens > 1000",
+		},
+		{
+			// an unknown name under the trace context still routes trace-level, so the
+			// aggregate validation rejects it with a targeted error instead of the span
+			// path failing on an unknown field.
+			name:   "unknown aggregate under trace context stays trace-level",
+			query:  "trace.not_an_aggregate > 1000",
+			having: "trace.not_an_aggregate > 1000",
+		},
+
+		{
+			// ANTLR token offsets are rune indices; slicing must not shift after a
+			// multi-byte char (this used to truncate 1000 → 100).
+			name:   "unicode value before the split",
+			query:  "service.name = 'héllo' AND completion_tokens > 1000",
+			span:   "service.name = 'héllo'",
+			having: "completion_tokens > 1000",
+		},
+
+		// --- top-level AND splits across the two buckets -------------------------
+		{
+			name:   "span AND agg",
+			query:  "service.name = 'x' AND completion_tokens > 1000",
+			span:   "service.name = 'x'",
+			having: "completion_tokens > 1000",
+		},
+		{
+			// order within a bucket is preserved; the two span atoms join with AND.
+			name:   "span AND span AND agg",
+			query:  "service.name = 'x' AND kind_string = 'Internal' AND completion_tokens > 1000",
+			span:   "service.name = 'x' AND kind_string = 'Internal'",
+			having: "completion_tokens > 1000",
+		},
+		{
+			// a parenthesized top-level AND still splits across the two buckets.
+			name:   "parenthesized span AND agg",
+			query:  "(service.name = 'x' AND completion_tokens > 1000)",
+			span:   "service.name = 'x'",
+			having: "completion_tokens > 1000",
+		},
+
+		// --- OR groups are re-wrapped in parens so a later AND-join can't invert
+		//     precedence (`a AND (b OR c)` must not flatten to `a AND b OR c`) ------
+		{
+			name:   "agg OR agg",
+			query:  "completion_tokens > 1000 OR span_count > 3",
+			having: "(completion_tokens > 1000 OR span_count > 3)",
+		},
+		{
+			name:  "span OR span",
+			query: "service.name = 'x' OR kind_string = 'Internal'",
+			span:  "(service.name = 'x' OR kind_string = 'Internal')",
+		},
+		{
+			name:  "span AND (span OR span)",
+			query: "service.name = 'x' AND (kind_string = 'Internal' OR kind_string = 'Client')",
+			span:  "service.name = 'x' AND (kind_string = 'Internal' OR kind_string = 'Client')",
+		},
+		{
+			name:   "agg AND (agg OR agg)",
+			query:  "prompt_tokens > 5 AND (completion_tokens > 1000 OR span_count > 3)",
+			having: "prompt_tokens > 5 AND (completion_tokens > 1000 OR span_count > 3)",
+		},
+		{
+			// the OR group routes to span, the trailing aggregate to having.
+			name:   "span AND (span OR span) AND agg",
+			query:  "a.b = 'x' AND (c.d = 'y' OR e.f = 'z') AND completion_tokens > 1000",
+			span:   "a.b = 'x' AND (c.d = 'y' OR e.f = 'z')",
+			having: "completion_tokens > 1000",
+		},
+
+		// --- a nested AND group flattens across the buckets (no spurious parens) --
+		{
+			name:   "(span AND agg) AND agg",
+			query:  "(service.name = 'x' AND completion_tokens > 1000) AND prompt_tokens > 5",
+			span:   "service.name = 'x'",
+			having: "completion_tokens > 1000 AND prompt_tokens > 5",
+		},
+
+		// --- NOT wrapping a single-class group is routed whole to that class ------
+		{
+			name:   "not agg",
+			query:  "NOT (completion_tokens > 1000)",
+			having: "NOT (completion_tokens > 1000)",
+		},
+		{
+			name:  "not span",
+			query: "NOT (service.name = 'x')",
+			span:  "NOT (service.name = 'x')",
+		},
+
+		// --- class-mixing is rejected in an OR group, a NOT group, or a nested OR -
+		{
+			name:    "agg OR span rejected",
+			query:   "completion_tokens > 1000 OR service.name = 'x'",
+			wantErr: true,
+		},
+		{
+			name:    "not mixed rejected",
+			query:   "NOT (completion_tokens > 1000 AND service.name = 'x')",
+			wantErr: true,
+		},
+		{
+			name:    "span AND (agg OR span) rejected",
+			query:   "service.name = 'x' AND (completion_tokens > 1000 OR kind_string = 'Client')",
+			wantErr: true,
+		},
+	}
+
+	for _, c := range cases {
+		t.Run(c.name, func(t *testing.T) {
+			span, having, err := SplitFilterForAggregates(c.query, agg)
+			if c.wantErr {
+				require.Error(t, err)
+				return
+			}
+			require.NoError(t, err)
+			require.Equal(t, c.span, span, "span part")
+			require.Equal(t, c.having, having, "having part")
+		})
+	}
+}

pkg/querybuilder/having_rewriter.go

@@ -18,6 +18,19 @@ func NewHavingExpressionRewriter() *HavingExpressionRewriter {
 	}
 }
 
+// Rewrite rewrites and validates a HAVING expression against a caller-supplied
+// column map (user-facing name -> SQL identifier/expression). Values are inlined, so
+// the result is a bare SQL boolean expression with no bound args. Used by callers
+// that project their own aggregate columns (e.g. the AI trace list) rather than the
+// query's Aggregations.
+func (r *HavingExpressionRewriter) Rewrite(expression string, columnMap map[string]string) (string, error) {
+	if len(strings.TrimSpace(expression)) == 0 {
+		return "", nil
+	}
+	r.columnMap = columnMap
+	return r.rewriteAndValidate(expression)
+}
+
 // RewriteForTraces rewrites and validates the HAVING expression for a traces query.
 func (r *HavingExpressionRewriter) RewriteForTraces(expression string, aggregations []qbtypes.TraceAggregation) (string, error) {
 	if len(strings.TrimSpace(expression)) == 0 {

pkg/querybuilder/key_resolution.go

@@ -1,6 +1,7 @@
 package querybuilder
 
 import (
+	"encoding/json"
 	"fmt"
 
 	"github.com/SigNoz/signoz/pkg/errors"
@@ -68,6 +69,100 @@ func NewKeyNotFoundError(name string) error {
 	return errors.NewInvalidInputf(errors.CodeInvalidInput, "key `%s` not found", name).WithUrl(KeyNotFoundDocURL)
 }
 
+// NewKeyNotFoundWarning is the warning surfaced when a referenced key is absent from
+// metadata and the query falls back to synthesized keys.
+func NewKeyNotFoundWarning(name string) string {
+	return fmt.Sprintf("key `%s` not found in metadata; querying the underlying data directly. If this is unexpected, check the key name for typos.", name)
+}
+
+// SynthesizeKeys builds the field keys to query when metadata has no match: a qualified
+// key is honored as-is; a bare key defaults to attribute context with the data type
+// inferred from the operand, or fanned out across string/number/bool without one.
+func SynthesizeKeys(field *telemetrytypes.TelemetryFieldKey, value any) []*telemetrytypes.TelemetryFieldKey {
+	base := *field
+	if base.FieldContext == telemetrytypes.FieldContextUnspecified {
+		base.FieldContext = telemetrytypes.FieldContextAttribute
+	}
+	// Resource values are strings; pin the type so operand coercion applies.
+	if base.FieldContext == telemetrytypes.FieldContextResource &&
+		base.FieldDataType == telemetrytypes.FieldDataTypeUnspecified {
+		base.FieldDataType = telemetrytypes.FieldDataTypeString
+	}
+
+	// A set data type needs only one synthesized key.
+	if base.FieldDataType != telemetrytypes.FieldDataTypeUnspecified {
+		clone := base
+		return []*telemetrytypes.TelemetryFieldKey{&clone}
+	}
+
+	dataTypes := inferDataTypesFromOperand(value)
+	keys := make([]*telemetrytypes.TelemetryFieldKey, 0, len(dataTypes))
+	for _, dt := range dataTypes {
+		clone := base
+		clone.FieldDataType = dt
+		keys = append(keys, &clone)
+	}
+	return keys
+}
+
+// allVariantDataTypes is the fanout used when no operand pins the type (exists / group by).
+func allVariantDataTypes() []telemetrytypes.FieldDataType {
+	return []telemetrytypes.FieldDataType{
+		telemetrytypes.FieldDataTypeString,
+		telemetrytypes.FieldDataTypeNumber,
+		telemetrytypes.FieldDataTypeBool,
+	}
+}
+
+// inferDataTypesFromOperand maps an operand value to the data type(s) to query. With no
+// operand or an unrecognized type it fans out across all variants.
+func inferDataTypesFromOperand(value any) []telemetrytypes.FieldDataType {
+	switch v := value.(type) {
+	case nil:
+		return allVariantDataTypes()
+	case string:
+		return []telemetrytypes.FieldDataType{telemetrytypes.FieldDataTypeString}
+	case bool:
+		return []telemetrytypes.FieldDataType{telemetrytypes.FieldDataTypeBool}
+	case float32, float64, int, int8, int16, int32, int64, uint, uint8, uint16, uint32, uint64, json.Number:
+		return []telemetrytypes.FieldDataType{telemetrytypes.FieldDataTypeNumber}
+	case []any:
+		return inferDataTypesFromList(v)
+	default:
+		return allVariantDataTypes()
+	}
+}
+
+// inferDataTypesFromList derives the distinct data types present in an `in` list, kept
+// in string/number/bool order; an empty or all-unknown list fans out.
+func inferDataTypesFromList(values []any) []telemetrytypes.FieldDataType {
+	var hasString, hasNumber, hasBool bool
+	for _, v := range values {
+		switch v.(type) {
+		case string:
+			hasString = true
+		case bool:
+			hasBool = true
+		case float32, float64, int, int8, int16, int32, int64, uint, uint8, uint16, uint32, uint64, json.Number:
+			hasNumber = true
+		}
+	}
+	var out []telemetrytypes.FieldDataType
+	if hasString {
+		out = append(out, telemetrytypes.FieldDataTypeString)
+	}
+	if hasNumber {
+		out = append(out, telemetrytypes.FieldDataTypeNumber)
+	}
+	if hasBool {
+		out = append(out, telemetrytypes.FieldDataTypeBool)
+	}
+	if len(out) == 0 {
+		return allVariantDataTypes()
+	}
+	return out
+}
+
 // NewFunctionUnsupportedError returns the error for a has/hasAny/hasAll/hasToken operator
 // on a builder that doesn't support it (logs body only), or nil for other operators.
 func NewFunctionUnsupportedError(operator qbtypes.FilterOperator) error {

pkg/querybuilder/key_resolution_test.go

@@ -0,0 +1,89 @@
+package querybuilder
+
+import (
+	"testing"
+
+	"github.com/SigNoz/signoz/pkg/types/telemetrytypes"
+	"github.com/stretchr/testify/assert"
+	"github.com/stretchr/testify/require"
+)
+
+func TestSynthesizeKeys(t *testing.T) {
+	tests := []struct {
+		name          string
+		field         telemetrytypes.TelemetryFieldKey
+		value         any
+		wantContexts  []telemetrytypes.FieldContext
+		wantDataTypes []telemetrytypes.FieldDataType
+	}{
+		{
+			name:          "bare key with string operand infers string",
+			field:         telemetrytypes.TelemetryFieldKey{Name: "error.type"},
+			value:         "timeout",
+			wantContexts:  []telemetrytypes.FieldContext{telemetrytypes.FieldContextAttribute},
+			wantDataTypes: []telemetrytypes.FieldDataType{telemetrytypes.FieldDataTypeString},
+		},
+		{
+			name:          "bare key with number operand infers number",
+			field:         telemetrytypes.TelemetryFieldKey{Name: "http.status"},
+			value:         float64(500),
+			wantContexts:  []telemetrytypes.FieldContext{telemetrytypes.FieldContextAttribute},
+			wantDataTypes: []telemetrytypes.FieldDataType{telemetrytypes.FieldDataTypeNumber},
+		},
+		{
+			name:          "bare key with bool operand infers bool",
+			field:         telemetrytypes.TelemetryFieldKey{Name: "sampled"},
+			value:         true,
+			wantContexts:  []telemetrytypes.FieldContext{telemetrytypes.FieldContextAttribute},
+			wantDataTypes: []telemetrytypes.FieldDataType{telemetrytypes.FieldDataTypeBool},
+		},
+		{
+			name:          "bare key with no operand fans out across variants",
+			field:         telemetrytypes.TelemetryFieldKey{Name: "exception.type"},
+			value:         nil,
+			wantContexts:  []telemetrytypes.FieldContext{telemetrytypes.FieldContextAttribute, telemetrytypes.FieldContextAttribute, telemetrytypes.FieldContextAttribute},
+			wantDataTypes: []telemetrytypes.FieldDataType{telemetrytypes.FieldDataTypeString, telemetrytypes.FieldDataTypeNumber, telemetrytypes.FieldDataTypeBool},
+		},
+		{
+			name:          "qualified data type honored as-is",
+			field:         telemetrytypes.TelemetryFieldKey{Name: "custom.key", FieldDataType: telemetrytypes.FieldDataTypeString},
+			value:         nil,
+			wantContexts:  []telemetrytypes.FieldContext{telemetrytypes.FieldContextAttribute},
+			wantDataTypes: []telemetrytypes.FieldDataType{telemetrytypes.FieldDataTypeString},
+		},
+		{
+			name:          "qualified resource context honored as single string key",
+			field:         telemetrytypes.TelemetryFieldKey{Name: "k8s.cluster.name", FieldContext: telemetrytypes.FieldContextResource},
+			value:         nil,
+			wantContexts:  []telemetrytypes.FieldContext{telemetrytypes.FieldContextResource},
+			wantDataTypes: []telemetrytypes.FieldDataType{telemetrytypes.FieldDataTypeString},
+		},
+		{
+			name:          "in list with homogeneous strings infers single string variant",
+			field:         telemetrytypes.TelemetryFieldKey{Name: "error.type"},
+			value:         []any{"a", "b"},
+			wantContexts:  []telemetrytypes.FieldContext{telemetrytypes.FieldContextAttribute},
+			wantDataTypes: []telemetrytypes.FieldDataType{telemetrytypes.FieldDataTypeString},
+		},
+		{
+			name:          "in list with mixed types fans out to present variants",
+			field:         telemetrytypes.TelemetryFieldKey{Name: "error.type"},
+			value:         []any{"a", float64(1)},
+			wantContexts:  []telemetrytypes.FieldContext{telemetrytypes.FieldContextAttribute, telemetrytypes.FieldContextAttribute},
+			wantDataTypes: []telemetrytypes.FieldDataType{telemetrytypes.FieldDataTypeString, telemetrytypes.FieldDataTypeNumber},
+		},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			field := tt.field
+			keys := SynthesizeKeys(&field, tt.value)
+			require.Len(t, keys, len(tt.wantDataTypes))
+			for i, k := range keys {
+				assert.Equal(t, tt.field.Name, k.Name, "name preserved")
+				assert.Equal(t, tt.wantContexts[i], k.FieldContext)
+				assert.Equal(t, tt.wantDataTypes[i], k.FieldDataType)
+			}
+		})
+	}
+}

pkg/querybuilder/where_clause_visitor.go

@@ -12,6 +12,7 @@ import (
 	grammar "github.com/SigNoz/signoz/pkg/parser/filterquery/grammar"
 	qbtypes "github.com/SigNoz/signoz/pkg/types/querybuildertypes/querybuildertypesv5"
 	"github.com/SigNoz/signoz/pkg/types/telemetrytypes"
+	"github.com/SigNoz/signoz/pkg/valuer"
 	"github.com/antlr4-go/antlr/v4"
 
 	sqlbuilder "github.com/huandu/go-sqlbuilder"
@@ -25,6 +26,7 @@ const stringMatchingOperatorDocURL = "https://signoz.io/docs/userguide/operators
 // to convert the parsed filter expressions into ClickHouse WHERE clause.
 type filterExpressionVisitor struct {
 	context            context.Context
+	orgID              valuer.UUID
 	fieldMapper        qbtypes.FieldMapper
 	conditionBuilder   qbtypes.ConditionBuilder
 	warnings           []string
@@ -45,6 +47,7 @@ type filterExpressionVisitor struct {
 
 type FilterExprVisitorOpts struct {
 	Context            context.Context
+	OrgID              valuer.UUID
 	Logger             *slog.Logger
 	FieldMapper        qbtypes.FieldMapper
 	ConditionBuilder   qbtypes.ConditionBuilder
@@ -62,6 +65,7 @@ type FilterExprVisitorOpts struct {
 func newFilterExpressionVisitor(opts FilterExprVisitorOpts) *filterExpressionVisitor {
 	return &filterExpressionVisitor{
 		context:            opts.Context,
+		orgID:              opts.OrgID,
 		fieldMapper:        opts.FieldMapper,
 		conditionBuilder:   opts.ConditionBuilder,
 		fieldKeys:          opts.FieldKeys,
@@ -365,12 +369,11 @@ func (v *filterExpressionVisitor) VisitPrimary(ctx *grammar.PrimaryContext) any
 // VisitComparison handles all comparison operators.
 func (v *filterExpressionVisitor) VisitComparison(ctx *grammar.ComparisonContext) any {
 	key := v.Visit(ctx.Key()).(*telemetrytypes.TelemetryFieldKey)
-	matching := matchingFieldKeys(key, v.fieldKeys)
+	matching := MatchingFieldKeys(key, v.fieldKeys)
 
-	// Skip resource filtering on the main table when a sub-query covers it. Resolve
-	// ambiguity first (resource+attribute defaults to resource), then drop resource
-	// matches; skip the term if nothing remains. Empty matches flow to the builder.
-	if v.skipResourceFilter && len(matching) > 0 {
+	// Skip resource filtering on the main table when a sub-query covers it; a resolving
+	// condition builder applies this itself in ConditionForKeys.
+	if _, resolvesKeys := v.conditionBuilder.(qbtypes.ResolvingConditionBuilder); v.skipResourceFilter && len(matching) > 0 && !resolvesKeys {
 		resolved, warning := ResolveKeys(key, matching)
 		// emit the ambiguity warning even when the term is skipped below
 		v.addWarnings([]string{warning}, len(matching) > 1)
@@ -731,7 +734,7 @@ func (v *filterExpressionVisitor) VisitFunctionCall(ctx *grammar.FunctionCallCon
 
 	value := params[1:]
 
-	conds, ok := v.buildConditions(key, matchingFieldKeys(key, v.fieldKeys), operator, value)
+	conds, ok := v.buildConditions(key, MatchingFieldKeys(key, v.fieldKeys), operator, value)
 	if !ok {
 		return ErrorConditionLiteral
 	}
@@ -811,10 +814,20 @@ func (v *filterExpressionVisitor) VisitKey(ctx *grammar.KeyContext) any {
 }
 
 // buildConditions invokes the condition builder for a filter term, folding its
-// warnings/errors into visitor state. It returns the conditions and false if an error
-// was recorded.
+// warnings/errors into visitor state; returns false if an error was recorded.
 func (v *filterExpressionVisitor) buildConditions(key *telemetrytypes.TelemetryFieldKey, matching []*telemetrytypes.TelemetryFieldKey, op qbtypes.FilterOperator, value any) ([]string, bool) {
-	conds, warns, err := v.conditionBuilder.ConditionFor(v.context, v.startNs, v.endNs, key, matching, op, value, v.builder)
+	var (
+		conds []string
+		warns []string
+		err   error
+	)
+	// A resolving condition builder owns key resolution, so hand it the raw key + full map;
+	// other signals use the pre-matched ConditionFor path.
+	if rcb, ok := v.conditionBuilder.(qbtypes.ResolvingConditionBuilder); ok {
+		conds, warns, err = rcb.ConditionForKeys(v.context, v.orgID, v.startNs, v.endNs, key, v.fieldKeys, qbtypes.ConditionBuilderOptions{SkipResourceFilter: v.skipResourceFilter}, op, value, v.builder)
+	} else {
+		conds, warns, err = v.conditionBuilder.ConditionFor(v.context, v.orgID, v.startNs, v.endNs, key, matching, op, value, v.builder)
+	}
 	if err != nil {
 		_, _, _, _, errURL, _ := errors.Unwrapb(err)
 		assignIfEmpty(&v.mainErrorURL, errURL)
@@ -870,10 +883,9 @@ func assignIfEmpty(s *string, value string) {
 	}
 }
 
-// matchingFieldKeys returns the field keys from the provided map that match the given
-// key, honoring any context/data type the user specified. It also resolves the case
-// where GetFieldKeyFromKeyText split a context off a name that legitimately contained it.
-func matchingFieldKeys(field *telemetrytypes.TelemetryFieldKey, fieldKeys map[string][]*telemetrytypes.TelemetryFieldKey) []*telemetrytypes.TelemetryFieldKey {
+// MatchingFieldKeys returns the field keys from the map that match the given key,
+// honoring any context/data type the user specified.
+func MatchingFieldKeys(field *telemetrytypes.TelemetryFieldKey, fieldKeys map[string][]*telemetrytypes.TelemetryFieldKey) []*telemetrytypes.TelemetryFieldKey {
 	fieldKeysForName := []*telemetrytypes.TelemetryFieldKey{}
 
 	// match by name; keep items whose context and data type match (unspecified matches any)

pkg/querybuilder/where_clause_visitor_test.go

@@ -10,6 +10,7 @@ import (
 	grammar "github.com/SigNoz/signoz/pkg/parser/filterquery/grammar"
 	qbtypes "github.com/SigNoz/signoz/pkg/types/querybuildertypes/querybuildertypesv5"
 	"github.com/SigNoz/signoz/pkg/types/telemetrytypes"
+	"github.com/SigNoz/signoz/pkg/valuer"
 	"github.com/antlr4-go/antlr/v4"
 	sqlbuilder "github.com/huandu/go-sqlbuilder"
 	"github.com/stretchr/testify/assert"
@@ -588,7 +589,7 @@ func TestVisitKey(t *testing.T) {
 			// VisitKey only parses; the condition builder matches, resolves ambiguity
 			// and decides not-found handling. Replay that here against the generic
 			// builder behavior (error unless the key is ignored).
-			matching := matchingFieldKeys(key, tt.fieldKeys)
+			matching := MatchingFieldKeys(key, tt.fieldKeys)
 			keys, warning := ResolveKeys(key, matching)
 
 			var gotErrors []string
@@ -746,6 +747,7 @@ type resourceConditionBuilder struct{}
 
 func (b *resourceConditionBuilder) ConditionFor(
 	_ context.Context,
+	_ valuer.UUID,
 	_ uint64,
 	_ uint64,
 	key *telemetrytypes.TelemetryFieldKey,
@@ -781,6 +783,7 @@ type conditionBuilder struct{}
 
 func (b *conditionBuilder) ConditionFor(
 	_ context.Context,
+	_ valuer.UUID,
 	_ uint64,
 	_ uint64,
 	key *telemetrytypes.TelemetryFieldKey,

pkg/signoz/provider.go

@@ -44,7 +44,6 @@ import (
 	"github.com/SigNoz/signoz/pkg/pprof/nooppprof"
 	"github.com/SigNoz/signoz/pkg/prometheus"
 	"github.com/SigNoz/signoz/pkg/prometheus/clickhouseprometheus"
-	"github.com/SigNoz/signoz/pkg/prometheus/clickhouseprometheusv2"
 	"github.com/SigNoz/signoz/pkg/querier"
 	"github.com/SigNoz/signoz/pkg/querier/signozquerier"
 	"github.com/SigNoz/signoz/pkg/sharder"
@@ -238,7 +237,6 @@ func NewTelemetryStoreProviderFactories() factory.NamedMap[factory.ProviderFacto
 func NewPrometheusProviderFactories(telemetryStore telemetrystore.TelemetryStore) factory.NamedMap[factory.ProviderFactory[prometheus.Prometheus, prometheus.Config]] {
 	return factory.MustNewNamedMap(
 		clickhouseprometheus.NewFactory(telemetryStore),
-		clickhouseprometheusv2.NewFactory(telemetryStore),
 	)
 }
 
@@ -280,9 +278,9 @@ func NewStatsReporterProviderFactories(aggregator statsreporter.Aggregator, orgG
 	)
 }
 
-func NewQuerierProviderFactories(telemetryStore telemetrystore.TelemetryStore, prometheus prometheus.Prometheus, promV2 *clickhouseprometheusv2.Provider, cache cache.Cache, flagger flagger.Flagger) factory.NamedMap[factory.ProviderFactory[querier.Querier, querier.Config]] {
+func NewQuerierProviderFactories(telemetryStore telemetrystore.TelemetryStore, prometheus prometheus.Prometheus, cache cache.Cache, flagger flagger.Flagger) factory.NamedMap[factory.ProviderFactory[querier.Querier, querier.Config]] {
 	return factory.MustNewNamedMap(
-		signozquerier.NewFactory(telemetryStore, prometheus, promV2, cache, flagger),
+		signozquerier.NewFactory(telemetryStore, prometheus, cache, flagger),
 	)
 }
 

pkg/signoz/signoz.go

@@ -39,7 +39,6 @@ import (
 	"github.com/SigNoz/signoz/pkg/modules/tag/impltag"
 	"github.com/SigNoz/signoz/pkg/modules/user/impluser"
 	"github.com/SigNoz/signoz/pkg/prometheus"
-	"github.com/SigNoz/signoz/pkg/prometheus/clickhouseprometheusv2"
 	"github.com/SigNoz/signoz/pkg/querier"
 	"github.com/SigNoz/signoz/pkg/queryparser"
 	"github.com/SigNoz/signoz/pkg/ruler"
@@ -242,11 +241,6 @@ func New(
 
 	retentionGetter := implretention.NewGetter(implretention.NewStore(sqlstore))
 
-	// promV2 is the clickhousev2 provider handed to the querier for shadow
-	// comparison and pinned serving (declared before the serving provider,
-	// whose variable shadows the package name below).
-	var promV2 *clickhouseprometheusv2.Provider
-
 	// Initialize prometheus from the available prometheus provider factories
 	prometheus, err := factory.NewProviderFromNamedMap(
 		ctx,
@@ -259,29 +253,12 @@ func New(
 		return nil, err
 	}
 
-	// With the default provider, also stand up the clickhousev2 provider for
-	// the querier: PromQL queries shadow-compare against it behind the
-	// use_prometheus_clickhouse_v2 flag (see pkg/querier/promql_shadow.go).
-	// It never serves by default. An explicit
-	// prometheus::provider: clickhousev2 makes v2 the serving provider
-	// outright, so there is nothing to compare against.
-	if config.Prometheus.Provider() == "clickhouse" {
-		v2Config := config.Prometheus
-		// The v2 engine only evaluates shadow and pinned queries; disable its
-		// active query tracker so two trackers never share a file.
-		v2Config.ActiveQueryTrackerConfig.Enabled = false
-		promV2, err = clickhouseprometheusv2.New(ctx, providerSettings, v2Config, telemetrystore)
-		if err != nil {
-			return nil, err
-		}
-	}
-
 	// Initialize querier from the available querier provider factories
 	querier, err := factory.NewProviderFromNamedMap(
 		ctx,
 		providerSettings,
 		config.Querier,
-		NewQuerierProviderFactories(telemetrystore, prometheus, promV2, cache, flagger),
+		NewQuerierProviderFactories(telemetrystore, prometheus, cache, flagger),
 		config.Querier.Provider(),
 	)
 	if err != nil {

pkg/telemetryai/field_mapper.go

@@ -0,0 +1,99 @@
+package telemetryai
+
+import (
+	"strings"
+
+	scopedtraces "github.com/SigNoz/signoz/pkg/telemetryscopedtraces"
+	qbtypes "github.com/SigNoz/signoz/pkg/types/querybuildertypes/querybuildertypesv5"
+	"github.com/SigNoz/signoz/pkg/types/telemetrytypes"
+)
+
+// genAIBaseConditionProvider: an AI trace has >=1 gen_ai LLM, tool, or agent span.
+type genAIBaseConditionProvider struct {
+	keys []string
+}
+
+var _ scopedtraces.BaseConditionProvider = (*genAIBaseConditionProvider)(nil)
+
+func newGenAIBaseConditionProvider() scopedtraces.BaseConditionProvider {
+	return &genAIBaseConditionProvider{
+		keys: []string{telemetrytypes.GenAIRequestModel, telemetrytypes.GenAIToolName, telemetrytypes.GenAIAgentName},
+	}
+}
+
+func (p *genAIBaseConditionProvider) FilterExpression() string {
+	parts := make([]string, 0, len(p.keys))
+	for _, k := range p.keys {
+		parts = append(parts, k+" EXISTS")
+	}
+	return strings.Join(parts, " OR ")
+}
+
+func (p *genAIBaseConditionProvider) FieldKeys() []*telemetrytypes.TelemetryFieldKey {
+	// Definitions come from GenAIFieldDefinitions so they can't drift from the
+	// canonical semconv keys; copy to take the address.
+	keys := make([]*telemetrytypes.TelemetryFieldKey, 0, len(p.keys))
+	for _, k := range p.keys {
+		def := telemetrytypes.GenAIFieldDefinitions[k]
+		keys = append(keys, &def)
+	}
+	return keys
+}
+
+// genAIColumnProvider adds AI/LLM per-trace metrics on top of the common columns.
+type genAIColumnProvider struct{}
+
+var _ scopedtraces.ColumnProvider = (*genAIColumnProvider)(nil)
+
+func newGenAIColumnProvider() scopedtraces.ColumnProvider {
+	return &genAIColumnProvider{}
+}
+
+func (genAIColumnProvider) Columns() []scopedtraces.TraceColumn {
+	defs := telemetrytypes.GenAIFieldDefinitions
+	reqModel := defs[telemetrytypes.GenAIRequestModel]
+	toolName := defs[telemetrytypes.GenAIToolName]
+	inTok := defs[telemetrytypes.GenAIUsageInputTokens]
+	outTok := defs[telemetrytypes.GenAIUsageOutputTokens]
+	cost := defs[telemetrytypes.SignozGenAITotalCost]
+	inMsg := defs[telemetrytypes.GenAIInputMessages]
+	outMsg := defs[telemetrytypes.GenAIOutputMessages]
+
+	str := telemetrytypes.FieldDataTypeString
+	return append(scopedtraces.CommonTraceColumns(),
+		// LLM calls only (request model present), not the full gate.
+		scopedtraces.TraceColumn{Alias: "llm_call_count", Orderable: true, Expr: scopedtraces.CountExists(&reqModel)},
+		scopedtraces.TraceColumn{Alias: "tool_call_count", Orderable: true, Expr: scopedtraces.CountExists(&toolName)},
+		scopedtraces.TraceColumn{Alias: "distinct_tool_count", Orderable: true, Expr: scopedtraces.UniqCount(&toolName, str)},
+		// tokens live only on LLM spans, so a plain sum needs no gate scoping.
+		scopedtraces.TraceColumn{Alias: "input_tokens", Orderable: true, Expr: scopedtraces.Reduce(scopedtraces.AggSum, &inTok)},
+		scopedtraces.TraceColumn{Alias: "output_tokens", Orderable: true, Expr: scopedtraces.Reduce(scopedtraces.AggSum, &outTok)},
+		scopedtraces.TraceColumn{Alias: "total_tokens", Orderable: true, Expr: scopedtraces.SumOfKeys(telemetrytypes.FieldDataTypeFloat64, &inTok, &outTok)},
+		// per-span cost attached by the SigNoz LLM pricing processor.
+		scopedtraces.TraceColumn{Alias: "estimated_cost_usd", Orderable: true, Expr: scopedtraces.Reduce(scopedtraces.AggSum, &cost)},
+		// slowest single LLM call in the trace.
+		scopedtraces.TraceColumn{Alias: "max_llm_latency_ns", Orderable: true, Expr: scopedtraces.ScopedToKeyColumn(scopedtraces.AggMax, scopedtraces.IntrinsicSpanKey("duration_nano"), &reqModel)},
+		// errors across the whole trace (any span), so display-only.
+		scopedtraces.TraceColumn{Alias: "error_count", Expr: scopedtraces.CondCount(scopedtraces.IntrinsicSpanKey("has_error"), qbtypes.FilterOperatorEqual, true)},
+		// timestamp of the last gen_ai span (LLM/tool/agent), hence gate-scoped.
+		scopedtraces.TraceColumn{Alias: "last_activity_time", Orderable: true, Expr: scopedtraces.ScopedReduce(scopedtraces.AggMax, scopedtraces.IntrinsicSpanKey("timestamp"))},
+		// previews: first call's input (the prompt), last call's output (the answer).
+		scopedtraces.TraceColumn{Alias: "input", SpanLevel: true, Expr: scopedtraces.PickBy(&inMsg, str, scopedtraces.IntrinsicSpanKey("timestamp"), scopedtraces.PickEarliest)},
+		scopedtraces.TraceColumn{Alias: "output", SpanLevel: true, Expr: scopedtraces.PickBy(&outMsg, str, scopedtraces.IntrinsicSpanKey("timestamp"), scopedtraces.PickLatest)},
+	)
+}
+
+func (genAIColumnProvider) DefaultOrderAlias() string { return "last_activity_time" }
+
+func (p genAIColumnProvider) AggregateAliases() []string {
+	// Derived from Columns() so a new column can't be forgotten; SpanLevel columns
+	// are filtered span-level, so skip them.
+	cols := p.Columns()
+	aliases := make([]string, 0, len(cols))
+	for _, c := range cols {
+		if !c.SpanLevel {
+			aliases = append(aliases, c.Alias)
+		}
+	}
+	return aliases
+}

pkg/telemetryai/statement_builder.go

@@ -0,0 +1,21 @@
+package telemetryai
+
+import (
+	"github.com/SigNoz/signoz/pkg/factory"
+	"github.com/SigNoz/signoz/pkg/flagger"
+	scopedtraces "github.com/SigNoz/signoz/pkg/telemetryscopedtraces"
+	qbtypes "github.com/SigNoz/signoz/pkg/types/querybuildertypes/querybuildertypesv5"
+	"github.com/SigNoz/signoz/pkg/types/telemetrytypes"
+)
+
+// NewAITraceStatementBuilder wires the generic scoped-trace builder with the gen_ai
+// gate and AI columns. This package holds only gen_ai domain knowledge; the query
+// topology lives in telemetryscopedtraces.
+func NewAITraceStatementBuilder(
+	settings factory.ProviderSettings,
+	metadataStore telemetrytypes.MetadataStore,
+	traceStmtBuilder qbtypes.StatementBuilder[qbtypes.TraceAggregation],
+	fl flagger.Flagger,
+) qbtypes.StatementBuilder[qbtypes.TraceAggregation] {
+	return scopedtraces.NewScopedTraceStatementBuilder(settings, metadataStore, newGenAIBaseConditionProvider(), newGenAIColumnProvider(), traceStmtBuilder, fl)
+}

pkg/telemetryai/statement_builder_test.go

@@ -0,0 +1,992 @@
+package telemetryai
+
+import (
+	"context"
+	"fmt"
+	"strings"
+	"testing"
+	"time"
+
+	"github.com/SigNoz/signoz/pkg/flagger/flaggertest"
+	"github.com/SigNoz/signoz/pkg/instrumentation/instrumentationtest"
+	"github.com/SigNoz/signoz/pkg/querybuilder"
+	scopedtraces "github.com/SigNoz/signoz/pkg/telemetryscopedtraces"
+	"github.com/SigNoz/signoz/pkg/telemetrytraces"
+	qbtypes "github.com/SigNoz/signoz/pkg/types/querybuildertypes/querybuildertypesv5"
+	"github.com/SigNoz/signoz/pkg/types/telemetrytypes"
+	"github.com/SigNoz/signoz/pkg/types/telemetrytypes/telemetrytypestest"
+	"github.com/SigNoz/signoz/pkg/valuer"
+	"github.com/stretchr/testify/require"
+)
+
+// otelKeysMap seeds the OpenTelemetry gen_ai semantic-convention keys the AI
+// queries reference, so the metadata-backed field resolution succeeds in tests.
+func otelKeysMap() map[string][]*telemetrytypes.TelemetryFieldKey {
+	strKey := func(name string) *telemetrytypes.TelemetryFieldKey {
+		return &telemetrytypes.TelemetryFieldKey{
+			Name:          name,
+			Signal:        telemetrytypes.SignalTraces,
+			FieldContext:  telemetrytypes.FieldContextAttribute,
+			FieldDataType: telemetrytypes.FieldDataTypeString,
+		}
+	}
+	numKey := func(name string) *telemetrytypes.TelemetryFieldKey {
+		return &telemetrytypes.TelemetryFieldKey{
+			Name:          name,
+			Signal:        telemetrytypes.SignalTraces,
+			FieldContext:  telemetrytypes.FieldContextAttribute,
+			FieldDataType: telemetrytypes.FieldDataTypeFloat64,
+		}
+	}
+
+	m := make(map[string][]*telemetrytypes.TelemetryFieldKey)
+
+	// gen_ai semconv keys sourced from the single source of truth, mirroring what the
+	// production metadata store surfaces via enrichWithGenAIKeys.
+	for name, def := range telemetrytypes.GenAIFieldDefinitions {
+		keyCopy := def
+		m[name] = []*telemetrytypes.TelemetryFieldKey{&keyCopy}
+	}
+
+	// Extra keys these tests reference that aren't gen_ai semconv definitions.
+	m["gen_ai.user.id"] = []*telemetrytypes.TelemetryFieldKey{strKey("gen_ai.user.id")}
+	m["_signoz.gen_ai.total_cost"] = []*telemetrytypes.TelemetryFieldKey{numKey("_signoz.gen_ai.total_cost")}
+	m["gen_ai.usage.cached_input_tokens"] = []*telemetrytypes.TelemetryFieldKey{numKey("gen_ai.usage.cached_input_tokens")}
+	m["has_error"] = []*telemetrytypes.TelemetryFieldKey{{
+		Name:          "has_error",
+		Signal:        telemetrytypes.SignalTraces,
+		FieldContext:  telemetrytypes.FieldContextSpan,
+		FieldDataType: telemetrytypes.FieldDataTypeBool,
+	}}
+	// service.name carries the resource-column evolutions like production metadata, so
+	// the rendered value expression prefers the JSON resource column over the legacy
+	// map (matching the standard traces builder tests).
+	m["service.name"] = []*telemetrytypes.TelemetryFieldKey{{
+		Name:          "service.name",
+		Signal:        telemetrytypes.SignalTraces,
+		FieldContext:  telemetrytypes.FieldContextResource,
+		FieldDataType: telemetrytypes.FieldDataTypeString,
+		Evolutions:    resourceEvolutions(),
+	}}
+	return m
+}
+
+// resourceEvolutions is the canonical resource-column timeline: the legacy
+// resources_string map at epoch 0 and the JSON resource column released inside the
+// test window (mirrors telemetrytraces' mockEvolutionData).
+func resourceEvolutions() []*telemetrytypes.EvolutionEntry {
+	return []*telemetrytypes.EvolutionEntry{
+		{
+			Signal:       telemetrytypes.SignalTraces,
+			ColumnName:   "resources_string",
+			ColumnType:   "Map(LowCardinality(String), String)",
+			FieldContext: telemetrytypes.FieldContextResource,
+			FieldName:    "__all__",
+			ReleaseTime:  time.Unix(0, 0),
+		},
+		{
+			Signal:       telemetrytypes.SignalTraces,
+			ColumnName:   "resource",
+			ColumnType:   "JSON()",
+			FieldContext: telemetrytypes.FieldContextResource,
+			FieldName:    "__all__",
+			ReleaseTime:  time.Date(2025, 5, 22, 22, 0, 0, 0, time.UTC),
+		},
+	}
+}
+
+// standard test window (ms), matching the traces builder tests.
+const (
+	testStartMs = uint64(1747947419000)
+	testEndMs   = uint64(1747983448000)
+)
+
+func newTestBuilder(t *testing.T) qbtypes.StatementBuilder[qbtypes.TraceAggregation] {
+	return newTestBuilderWithKeys(t, otelKeysMap())
+}
+
+// newTestBuilderWithKeys mirrors the production wiring in signozquerier's provider.
+// The gen_ai keys are seeded via keysMap here; in production the metadata store
+// surfaces them itself (enrichWithGenAIKeys).
+func newTestBuilderWithKeys(t *testing.T, keysMap map[string][]*telemetrytypes.TelemetryFieldKey) qbtypes.StatementBuilder[qbtypes.TraceAggregation] {
+	t.Helper()
+	settings := instrumentationtest.New().ToProviderSettings()
+	fm := telemetrytraces.NewFieldMapper()
+	cb := telemetrytraces.NewConditionBuilder(fm)
+	mockMetadataStore := telemetrytypestest.NewMockMetadataStore()
+	mockMetadataStore.KeysMap = keysMap
+	fl := flaggertest.New(t)
+	// In production the metadata store enriches gen_ai keys (enrichWithGenAIKeys);
+	// here the mock is seeded directly via keysMap.
+	metadataStore := telemetrytypes.MetadataStore(mockMetadataStore)
+	rewriter := querybuilder.NewAggExprRewriter(settings, nil, fm, cb, nil, fl)
+	traceStmtBuilder := telemetrytraces.NewTraceQueryStatementBuilder(
+		settings,
+		metadataStore,
+		fm,
+		cb,
+		rewriter,
+		nil,
+		fl,
+		false,
+		100000,
+	)
+	return NewAITraceStatementBuilder(
+		settings,
+		metadataStore,
+		traceStmtBuilder,
+		fl,
+	)
+}
+
+// ---------------------------------------------------------------------------
+// Full-query golden tests
+//
+// Each pins the WHOLE generated statement, with bound args inlined into the `?`
+// placeholders, as ONE self-contained literal — so a failure diff shows the entire
+// query and the expected SQL can be copied straight into a ClickHouse client. The
+// `want` strings are formatted for readability; the comparison is whitespace- and
+// backtick-insensitive (see normalizeSQL), so only the SQL tokens themselves matter.
+//
+// The four trace-list goldens cover the corners of how `matched` is assembled —
+// {no span filter, span filter} × {no aggregate filter, aggregate filter} — plus a
+// mixed filter + multi-key order, plus the delegated span list. Note `matched` selects
+// only the aggregates ORDER BY / HAVING reference; the rest appear only in enrichment.
+//
+// Run `go test ./pkg/telemetryai/ -run TestBuild_FullSQL -v` to also print each query.
+// ---------------------------------------------------------------------------
+
+// renderSQL substitutes bound args into the `?` placeholders so the whole statement
+// reads as one literal SQL string.
+func renderSQL(t *testing.T, stmt *qbtypes.Statement) string {
+	t.Helper()
+	var b strings.Builder
+	argi := 0
+	for i := 0; i < len(stmt.Query); i++ {
+		if stmt.Query[i] == '?' {
+			require.Less(t, argi, len(stmt.Args), "more ? than args in query")
+			b.WriteString(formatArg(stmt.Args[argi]))
+			argi++
+			continue
+		}
+		b.WriteByte(stmt.Query[i])
+	}
+	require.Equal(t, len(stmt.Args), argi, "arg count does not match number of placeholders")
+	return b.String()
+}
+
+func formatArg(a any) string {
+	if s, ok := a.(string); ok {
+		return "'" + s + "'"
+	}
+	return fmt.Sprintf("%v", a)
+}
+
+// normalizeSQL makes the comparison insensitive to formatting: it drops identifier
+// backticks, collapses whitespace runs to a single space, and removes spaces directly
+// inside parentheses. This lets the golden strings be freely indented/wrapped (and
+// written as Go raw literals, which cannot contain backticks) — only the SQL tokens
+// and their order matter.
+func normalizeSQL(s string) string {
+	s = strings.Join(strings.Fields(strings.ReplaceAll(s, "`", "")), " ")
+	s = strings.ReplaceAll(s, "( ", "(")
+	s = strings.ReplaceAll(s, " )", ")")
+	return s
+}
+
+func requireSQLEqual(t *testing.T, want string, stmt *qbtypes.Statement) {
+	t.Helper()
+	got := renderSQL(t, stmt)
+	t.Logf("\n%s", got)
+	require.Equal(t, normalizeSQL(want), normalizeSQL(got))
+}
+
+// No filter: matched selects only the default order key (last_activity_time), WHERE is
+// just window + gate mask, no HAVING.
+func TestBuild_FullSQL_TraceList_NoFilter(t *testing.T) {
+	b := newTestBuilder(t)
+	stmt, err := b.Build(context.Background(), valuer.UUID{}, testStartMs, testEndMs, qbtypes.RequestTypeTrace,
+		qbtypes.QueryBuilderQuery[qbtypes.TraceAggregation]{
+			Signal: telemetrytypes.SignalTraces, Source: telemetrytypes.SourceAI, Limit: 20,
+		}, nil)
+	require.NoError(t, err)
+
+	requireSQLEqual(t, `
+WITH matched AS (
+    SELECT trace_id,
+        maxIf(timestamp, (mapContains(attributes_string, 'gen_ai.request.model') = true OR mapContains(attributes_string, 'gen_ai.tool.name') = true OR mapContains(attributes_string, 'gen_ai.agent.name') = true)) AS last_activity_time
+    FROM signoz_traces.distributed_signoz_index_v3
+    WHERE timestamp >= '1747947419000000000'
+      AND timestamp < '1747983448000000000'
+      AND ts_bucket_start >= 1747945619
+      AND ts_bucket_start <= 1747983448
+      AND ((mapContains(attributes_string, 'gen_ai.request.model') = true OR mapContains(attributes_string, 'gen_ai.tool.name') = true OR mapContains(attributes_string, 'gen_ai.agent.name') = true))
+    GROUP BY trace_id
+    ORDER BY last_activity_time DESC, trace_id DESC
+    LIMIT 20
+),
+ranked AS (
+    SELECT trace_id, min(start) AS t_start, max(end) AS t_end
+    FROM signoz_traces.distributed_trace_summary
+    WHERE trace_id GLOBAL IN (SELECT trace_id FROM matched)
+      AND end >= fromUnixTimestamp64Nano(1747947419000000000)
+      AND start < fromUnixTimestamp64Nano(1747983448000000000)
+    GROUP BY trace_id
+),
+buckets AS (
+    SELECT DISTINCT b AS ts_bucket
+    FROM ranked
+    ARRAY JOIN range(toUInt64(intDiv(toUnixTimestamp(t_start), 1800) * 1800 - 1800), toUInt64(intDiv(toUnixTimestamp(t_end), 1800) * 1800 + 1800), 1800) AS b
+)
+SELECT trace_id,
+    min(timestamp) AS start_time,
+    max(timestamp) AS end_time,
+    (max(toUnixTimestamp64Nano(timestamp) + duration_nano) - min(toUnixTimestamp64Nano(timestamp))) AS trace_duration_nano,
+    count() AS span_count,
+    anyIf(name, parent_span_id = '') AS root_span_name,
+    any(multiIf(multiIf(resource.service.name IS NOT NULL, resource.service.name::String, mapContains(resources_string, 'service.name'), resources_string['service.name'], NULL) IS NOT NULL, multiIf(resource.service.name IS NOT NULL, resource.service.name::String, mapContains(resources_string, 'service.name'), resources_string['service.name'], NULL), NULL)) AS service.name,
+    countIf(mapContains(attributes_string, 'gen_ai.request.model') = true) AS llm_call_count,
+    countIf(mapContains(attributes_string, 'gen_ai.tool.name') = true) AS tool_call_count,
+    uniqIf(multiIf(mapContains(attributes_string, 'gen_ai.tool.name') = true, attributes_string['gen_ai.tool.name'], NULL), mapContains(attributes_string, 'gen_ai.tool.name') = true) AS distinct_tool_count,
+    sum(multiIf(mapContains(attributes_number, 'gen_ai.usage.input_tokens') = true, toFloat64(attributes_number['gen_ai.usage.input_tokens']), NULL)) AS input_tokens,
+    sum(multiIf(mapContains(attributes_number, 'gen_ai.usage.output_tokens') = true, toFloat64(attributes_number['gen_ai.usage.output_tokens']), NULL)) AS output_tokens,
+    coalesce(sum(multiIf(mapContains(attributes_number, 'gen_ai.usage.input_tokens') = true, toFloat64(attributes_number['gen_ai.usage.input_tokens']), NULL)), 0) + coalesce(sum(multiIf(mapContains(attributes_number, 'gen_ai.usage.output_tokens') = true, toFloat64(attributes_number['gen_ai.usage.output_tokens']), NULL)), 0) AS total_tokens,
+    sum(multiIf(mapContains(attributes_number, '_signoz.gen_ai.total_cost') = true, toFloat64(attributes_number['_signoz.gen_ai.total_cost']), NULL)) AS estimated_cost_usd,
+    maxIf(duration_nano, mapContains(attributes_string, 'gen_ai.request.model') = true) AS max_llm_latency_ns,
+    countIf(has_error = true) AS error_count,
+    maxIf(timestamp, (mapContains(attributes_string, 'gen_ai.request.model') = true OR mapContains(attributes_string, 'gen_ai.tool.name') = true OR mapContains(attributes_string, 'gen_ai.agent.name') = true)) AS last_activity_time,
+    argMinIf(multiIf(mapContains(attributes_string, 'gen_ai.input.messages') = true, attributes_string['gen_ai.input.messages'], NULL), timestamp, mapContains(attributes_string, 'gen_ai.input.messages') = true) AS input,
+    argMaxIf(multiIf(mapContains(attributes_string, 'gen_ai.output.messages') = true, attributes_string['gen_ai.output.messages'], NULL), timestamp, mapContains(attributes_string, 'gen_ai.output.messages') = true) AS output
+FROM signoz_traces.distributed_signoz_index_v3
+WHERE ts_bucket_start GLOBAL IN (SELECT ts_bucket FROM buckets)
+  AND trace_id GLOBAL IN (SELECT trace_id FROM ranked)
+GROUP BY trace_id
+ORDER BY last_activity_time DESC, trace_id DESC
+SETTINGS distributed_product_mode='allow', max_memory_usage=10000000000
+`, stmt)
+}
+
+// Promotion: a materialized gen_ai attribute must resolve to its materialized column
+// everywhere it appears — gate mask, countIf/scoped existence, and value columns —
+// while un-promoted attributes stay in the attributes map, so one query mixes both
+// forms. Here gen_ai.request.model and gen_ai.usage.input_tokens are materialized:
+// the gate/llm_call_count/max_llm_latency use `..._exists`, input_tokens/total_tokens
+// use the materialized value column, and tool/output_tokens/cost/messages stay in the map.
+func TestBuild_FullSQL_TraceList_MaterializedColumns(t *testing.T) {
+	keys := otelKeysMap()
+	for _, name := range []string{"gen_ai.request.model", "gen_ai.usage.input_tokens"} {
+		for _, k := range keys[name] {
+			k.Materialized = true
+		}
+	}
+	b := newTestBuilderWithKeys(t, keys)
+	stmt, err := b.Build(context.Background(), valuer.UUID{}, testStartMs, testEndMs, qbtypes.RequestTypeTrace,
+		qbtypes.QueryBuilderQuery[qbtypes.TraceAggregation]{
+			Signal: telemetrytypes.SignalTraces, Source: telemetrytypes.SourceAI, Limit: 20,
+		}, nil)
+	require.NoError(t, err)
+
+	requireSQLEqual(t, `
+WITH matched AS (
+    SELECT trace_id,
+        maxIf(timestamp, (attribute_string_gen_ai$$request$$model_exists = true OR mapContains(attributes_string, 'gen_ai.tool.name') = true OR mapContains(attributes_string, 'gen_ai.agent.name') = true)) AS last_activity_time
+    FROM signoz_traces.distributed_signoz_index_v3
+    WHERE timestamp >= '1747947419000000000'
+      AND timestamp < '1747983448000000000'
+      AND ts_bucket_start >= 1747945619
+      AND ts_bucket_start <= 1747983448
+      AND ((attribute_string_gen_ai$$request$$model_exists = true OR mapContains(attributes_string, 'gen_ai.tool.name') = true OR mapContains(attributes_string, 'gen_ai.agent.name') = true))
+    GROUP BY trace_id
+    ORDER BY last_activity_time DESC, trace_id DESC
+    LIMIT 20
+),
+ranked AS (
+    SELECT trace_id, min(start) AS t_start, max(end) AS t_end
+    FROM signoz_traces.distributed_trace_summary
+    WHERE trace_id GLOBAL IN (SELECT trace_id FROM matched)
+      AND end >= fromUnixTimestamp64Nano(1747947419000000000)
+      AND start < fromUnixTimestamp64Nano(1747983448000000000)
+    GROUP BY trace_id
+),
+buckets AS (
+    SELECT DISTINCT b AS ts_bucket
+    FROM ranked
+    ARRAY JOIN range(toUInt64(intDiv(toUnixTimestamp(t_start), 1800) * 1800 - 1800), toUInt64(intDiv(toUnixTimestamp(t_end), 1800) * 1800 + 1800), 1800) AS b
+)
+SELECT trace_id,
+    min(timestamp) AS start_time,
+    max(timestamp) AS end_time,
+    (max(toUnixTimestamp64Nano(timestamp) + duration_nano) - min(toUnixTimestamp64Nano(timestamp))) AS trace_duration_nano,
+    count() AS span_count,
+    anyIf(name, parent_span_id = '') AS root_span_name,
+    any(multiIf(multiIf(resource.service.name IS NOT NULL, resource.service.name::String, mapContains(resources_string, 'service.name'), resources_string['service.name'], NULL) IS NOT NULL, multiIf(resource.service.name IS NOT NULL, resource.service.name::String, mapContains(resources_string, 'service.name'), resources_string['service.name'], NULL), NULL)) AS service.name,
+    countIf(attribute_string_gen_ai$$request$$model_exists = true) AS llm_call_count,
+    countIf(mapContains(attributes_string, 'gen_ai.tool.name') = true) AS tool_call_count,
+    uniqIf(multiIf(mapContains(attributes_string, 'gen_ai.tool.name') = true, attributes_string['gen_ai.tool.name'], NULL), mapContains(attributes_string, 'gen_ai.tool.name') = true) AS distinct_tool_count,
+    sum(multiIf(attribute_number_gen_ai$$usage$$input_tokens_exists = true, toFloat64(attribute_number_gen_ai$$usage$$input_tokens), NULL)) AS input_tokens,
+    sum(multiIf(mapContains(attributes_number, 'gen_ai.usage.output_tokens') = true, toFloat64(attributes_number['gen_ai.usage.output_tokens']), NULL)) AS output_tokens,
+    coalesce(sum(multiIf(attribute_number_gen_ai$$usage$$input_tokens_exists = true, toFloat64(attribute_number_gen_ai$$usage$$input_tokens), NULL)), 0) + coalesce(sum(multiIf(mapContains(attributes_number, 'gen_ai.usage.output_tokens') = true, toFloat64(attributes_number['gen_ai.usage.output_tokens']), NULL)), 0) AS total_tokens,
+    sum(multiIf(mapContains(attributes_number, '_signoz.gen_ai.total_cost') = true, toFloat64(attributes_number['_signoz.gen_ai.total_cost']), NULL)) AS estimated_cost_usd,
+    maxIf(duration_nano, attribute_string_gen_ai$$request$$model_exists = true) AS max_llm_latency_ns,
+    countIf(has_error = true) AS error_count,
+    maxIf(timestamp, (attribute_string_gen_ai$$request$$model_exists = true OR mapContains(attributes_string, 'gen_ai.tool.name') = true OR mapContains(attributes_string, 'gen_ai.agent.name') = true)) AS last_activity_time,
+    argMinIf(multiIf(mapContains(attributes_string, 'gen_ai.input.messages') = true, attributes_string['gen_ai.input.messages'], NULL), timestamp, mapContains(attributes_string, 'gen_ai.input.messages') = true) AS input,
+    argMaxIf(multiIf(mapContains(attributes_string, 'gen_ai.output.messages') = true, attributes_string['gen_ai.output.messages'], NULL), timestamp, mapContains(attributes_string, 'gen_ai.output.messages') = true) AS output
+FROM signoz_traces.distributed_signoz_index_v3
+WHERE ts_bucket_start GLOBAL IN (SELECT ts_bucket FROM buckets)
+  AND trace_id GLOBAL IN (SELECT trace_id FROM ranked)
+GROUP BY trace_id
+ORDER BY last_activity_time DESC, trace_id DESC
+SETTINGS distributed_product_mode='allow', max_memory_usage=10000000000
+`, stmt)
+}
+
+// Span-level AND trace-level filter, order by the aggregate, pagination. matched selects
+// only output_tokens (the sole aggregate referenced by both ORDER BY and HAVING) — not
+// input_tokens/llm_call_count/last_activity_time. The span predicate widens the WHERE
+// prune and becomes a countIf(...) > 0 existence check alongside the gate countIf.
+func TestBuild_FullSQL_TraceList_SpanAndTraceFilter(t *testing.T) {
+	b := newTestBuilder(t)
+	stmt, err := b.Build(context.Background(), valuer.UUID{}, testStartMs, testEndMs, qbtypes.RequestTypeTrace,
+		qbtypes.QueryBuilderQuery[qbtypes.TraceAggregation]{
+			Signal: telemetrytypes.SignalTraces, Source: telemetrytypes.SourceAI,
+			Filter: &qbtypes.Filter{Expression: "gen_ai.request.model = 'gpt-4o-mini' AND output_tokens > 1000"},
+			Order:  []qbtypes.OrderBy{{Key: qbtypes.OrderByKey{TelemetryFieldKey: telemetrytypes.TelemetryFieldKey{Name: "output_tokens"}}, Direction: qbtypes.OrderDirectionDesc}},
+			Limit:  10, Offset: 30,
+		}, nil)
+	require.NoError(t, err)
+
+	requireSQLEqual(t, `
+WITH matched AS (
+    SELECT trace_id,
+        sum(multiIf(mapContains(attributes_number, 'gen_ai.usage.output_tokens') = true, toFloat64(attributes_number['gen_ai.usage.output_tokens']), NULL)) AS output_tokens
+    FROM signoz_traces.distributed_signoz_index_v3
+    WHERE timestamp >= '1747947419000000000'
+      AND timestamp < '1747983448000000000'
+      AND ts_bucket_start >= 1747945619
+      AND ts_bucket_start <= 1747983448
+      AND ((mapContains(attributes_string, 'gen_ai.request.model') = true OR mapContains(attributes_string, 'gen_ai.tool.name') = true OR mapContains(attributes_string, 'gen_ai.agent.name') = true)
+        OR (attributes_string['gen_ai.request.model'] = 'gpt-4o-mini' AND mapContains(attributes_string, 'gen_ai.request.model') = true))
+    GROUP BY trace_id
+    HAVING countIf((mapContains(attributes_string, 'gen_ai.request.model') = true OR mapContains(attributes_string, 'gen_ai.tool.name') = true OR mapContains(attributes_string, 'gen_ai.agent.name') = true)) > 0
+        AND countIf((attributes_string['gen_ai.request.model'] = 'gpt-4o-mini' AND mapContains(attributes_string, 'gen_ai.request.model') = true)) > 0
+        AND output_tokens > 1000
+    ORDER BY output_tokens DESC, trace_id DESC
+    LIMIT 10 OFFSET 30
+),
+ranked AS (
+    SELECT trace_id, min(start) AS t_start, max(end) AS t_end
+    FROM signoz_traces.distributed_trace_summary
+    WHERE trace_id GLOBAL IN (SELECT trace_id FROM matched)
+      AND end >= fromUnixTimestamp64Nano(1747947419000000000)
+      AND start < fromUnixTimestamp64Nano(1747983448000000000)
+    GROUP BY trace_id
+),
+buckets AS (
+    SELECT DISTINCT b AS ts_bucket
+    FROM ranked
+    ARRAY JOIN range(toUInt64(intDiv(toUnixTimestamp(t_start), 1800) * 1800 - 1800), toUInt64(intDiv(toUnixTimestamp(t_end), 1800) * 1800 + 1800), 1800) AS b
+)
+SELECT trace_id,
+    min(timestamp) AS start_time,
+    max(timestamp) AS end_time,
+    (max(toUnixTimestamp64Nano(timestamp) + duration_nano) - min(toUnixTimestamp64Nano(timestamp))) AS trace_duration_nano,
+    count() AS span_count,
+    anyIf(name, parent_span_id = '') AS root_span_name,
+    any(multiIf(multiIf(resource.service.name IS NOT NULL, resource.service.name::String, mapContains(resources_string, 'service.name'), resources_string['service.name'], NULL) IS NOT NULL, multiIf(resource.service.name IS NOT NULL, resource.service.name::String, mapContains(resources_string, 'service.name'), resources_string['service.name'], NULL), NULL)) AS service.name,
+    countIf(mapContains(attributes_string, 'gen_ai.request.model') = true) AS llm_call_count,
+    countIf(mapContains(attributes_string, 'gen_ai.tool.name') = true) AS tool_call_count,
+    uniqIf(multiIf(mapContains(attributes_string, 'gen_ai.tool.name') = true, attributes_string['gen_ai.tool.name'], NULL), mapContains(attributes_string, 'gen_ai.tool.name') = true) AS distinct_tool_count,
+    sum(multiIf(mapContains(attributes_number, 'gen_ai.usage.input_tokens') = true, toFloat64(attributes_number['gen_ai.usage.input_tokens']), NULL)) AS input_tokens,
+    sum(multiIf(mapContains(attributes_number, 'gen_ai.usage.output_tokens') = true, toFloat64(attributes_number['gen_ai.usage.output_tokens']), NULL)) AS output_tokens,
+    coalesce(sum(multiIf(mapContains(attributes_number, 'gen_ai.usage.input_tokens') = true, toFloat64(attributes_number['gen_ai.usage.input_tokens']), NULL)), 0) + coalesce(sum(multiIf(mapContains(attributes_number, 'gen_ai.usage.output_tokens') = true, toFloat64(attributes_number['gen_ai.usage.output_tokens']), NULL)), 0) AS total_tokens,
+    sum(multiIf(mapContains(attributes_number, '_signoz.gen_ai.total_cost') = true, toFloat64(attributes_number['_signoz.gen_ai.total_cost']), NULL)) AS estimated_cost_usd,
+    maxIf(duration_nano, mapContains(attributes_string, 'gen_ai.request.model') = true) AS max_llm_latency_ns,
+    countIf(has_error = true) AS error_count,
+    maxIf(timestamp, (mapContains(attributes_string, 'gen_ai.request.model') = true OR mapContains(attributes_string, 'gen_ai.tool.name') = true OR mapContains(attributes_string, 'gen_ai.agent.name') = true)) AS last_activity_time,
+    argMinIf(multiIf(mapContains(attributes_string, 'gen_ai.input.messages') = true, attributes_string['gen_ai.input.messages'], NULL), timestamp, mapContains(attributes_string, 'gen_ai.input.messages') = true) AS input,
+    argMaxIf(multiIf(mapContains(attributes_string, 'gen_ai.output.messages') = true, attributes_string['gen_ai.output.messages'], NULL), timestamp, mapContains(attributes_string, 'gen_ai.output.messages') = true) AS output
+FROM signoz_traces.distributed_signoz_index_v3
+WHERE ts_bucket_start GLOBAL IN (SELECT ts_bucket FROM buckets)
+  AND trace_id GLOBAL IN (SELECT trace_id FROM ranked)
+GROUP BY trace_id
+ORDER BY output_tokens DESC, trace_id DESC
+SETTINGS distributed_product_mode='allow', max_memory_usage=10000000000
+`, stmt)
+}
+
+// Aggregate-only filter (no span filter). WHERE prune is NOT widened, there is no
+// gate/span countIf, just the aggregate HAVING. `trace.output_tokens` rewrites to the
+// output_tokens alias. matched selects output_tokens (HAVING) + last_activity_time (default order).
+func TestBuild_FullSQL_TraceList_AggregateFilterOnly(t *testing.T) {
+	b := newTestBuilder(t)
+	stmt, err := b.Build(context.Background(), valuer.UUID{}, testStartMs, testEndMs, qbtypes.RequestTypeTrace,
+		qbtypes.QueryBuilderQuery[qbtypes.TraceAggregation]{
+			Signal: telemetrytypes.SignalTraces, Source: telemetrytypes.SourceAI,
+			Filter: &qbtypes.Filter{Expression: "trace.output_tokens > 1000"},
+			Limit:  20,
+		}, nil)
+	require.NoError(t, err)
+
+	requireSQLEqual(t, `
+WITH matched AS (
+    SELECT trace_id,
+        sum(multiIf(mapContains(attributes_number, 'gen_ai.usage.output_tokens') = true, toFloat64(attributes_number['gen_ai.usage.output_tokens']), NULL)) AS output_tokens,
+        maxIf(timestamp, (mapContains(attributes_string, 'gen_ai.request.model') = true OR mapContains(attributes_string, 'gen_ai.tool.name') = true OR mapContains(attributes_string, 'gen_ai.agent.name') = true)) AS last_activity_time
+    FROM signoz_traces.distributed_signoz_index_v3
+    WHERE timestamp >= '1747947419000000000'
+      AND timestamp < '1747983448000000000'
+      AND ts_bucket_start >= 1747945619
+      AND ts_bucket_start <= 1747983448
+      AND ((mapContains(attributes_string, 'gen_ai.request.model') = true OR mapContains(attributes_string, 'gen_ai.tool.name') = true OR mapContains(attributes_string, 'gen_ai.agent.name') = true))
+    GROUP BY trace_id
+    HAVING output_tokens > 1000
+    ORDER BY last_activity_time DESC, trace_id DESC
+    LIMIT 20
+),
+ranked AS (
+    SELECT trace_id, min(start) AS t_start, max(end) AS t_end
+    FROM signoz_traces.distributed_trace_summary
+    WHERE trace_id GLOBAL IN (SELECT trace_id FROM matched)
+      AND end >= fromUnixTimestamp64Nano(1747947419000000000)
+      AND start < fromUnixTimestamp64Nano(1747983448000000000)
+    GROUP BY trace_id
+),
+buckets AS (
+    SELECT DISTINCT b AS ts_bucket
+    FROM ranked
+    ARRAY JOIN range(toUInt64(intDiv(toUnixTimestamp(t_start), 1800) * 1800 - 1800), toUInt64(intDiv(toUnixTimestamp(t_end), 1800) * 1800 + 1800), 1800) AS b
+)
+SELECT trace_id,
+    min(timestamp) AS start_time,
+    max(timestamp) AS end_time,
+    (max(toUnixTimestamp64Nano(timestamp) + duration_nano) - min(toUnixTimestamp64Nano(timestamp))) AS trace_duration_nano,
+    count() AS span_count,
+    anyIf(name, parent_span_id = '') AS root_span_name,
+    any(multiIf(multiIf(resource.service.name IS NOT NULL, resource.service.name::String, mapContains(resources_string, 'service.name'), resources_string['service.name'], NULL) IS NOT NULL, multiIf(resource.service.name IS NOT NULL, resource.service.name::String, mapContains(resources_string, 'service.name'), resources_string['service.name'], NULL), NULL)) AS service.name,
+    countIf(mapContains(attributes_string, 'gen_ai.request.model') = true) AS llm_call_count,
+    countIf(mapContains(attributes_string, 'gen_ai.tool.name') = true) AS tool_call_count,
+    uniqIf(multiIf(mapContains(attributes_string, 'gen_ai.tool.name') = true, attributes_string['gen_ai.tool.name'], NULL), mapContains(attributes_string, 'gen_ai.tool.name') = true) AS distinct_tool_count,
+    sum(multiIf(mapContains(attributes_number, 'gen_ai.usage.input_tokens') = true, toFloat64(attributes_number['gen_ai.usage.input_tokens']), NULL)) AS input_tokens,
+    sum(multiIf(mapContains(attributes_number, 'gen_ai.usage.output_tokens') = true, toFloat64(attributes_number['gen_ai.usage.output_tokens']), NULL)) AS output_tokens,
+    coalesce(sum(multiIf(mapContains(attributes_number, 'gen_ai.usage.input_tokens') = true, toFloat64(attributes_number['gen_ai.usage.input_tokens']), NULL)), 0) + coalesce(sum(multiIf(mapContains(attributes_number, 'gen_ai.usage.output_tokens') = true, toFloat64(attributes_number['gen_ai.usage.output_tokens']), NULL)), 0) AS total_tokens,
+    sum(multiIf(mapContains(attributes_number, '_signoz.gen_ai.total_cost') = true, toFloat64(attributes_number['_signoz.gen_ai.total_cost']), NULL)) AS estimated_cost_usd,
+    maxIf(duration_nano, mapContains(attributes_string, 'gen_ai.request.model') = true) AS max_llm_latency_ns,
+    countIf(has_error = true) AS error_count,
+    maxIf(timestamp, (mapContains(attributes_string, 'gen_ai.request.model') = true OR mapContains(attributes_string, 'gen_ai.tool.name') = true OR mapContains(attributes_string, 'gen_ai.agent.name') = true)) AS last_activity_time,
+    argMinIf(multiIf(mapContains(attributes_string, 'gen_ai.input.messages') = true, attributes_string['gen_ai.input.messages'], NULL), timestamp, mapContains(attributes_string, 'gen_ai.input.messages') = true) AS input,
+    argMaxIf(multiIf(mapContains(attributes_string, 'gen_ai.output.messages') = true, attributes_string['gen_ai.output.messages'], NULL), timestamp, mapContains(attributes_string, 'gen_ai.output.messages') = true) AS output
+FROM signoz_traces.distributed_signoz_index_v3
+WHERE ts_bucket_start GLOBAL IN (SELECT ts_bucket FROM buckets)
+  AND trace_id GLOBAL IN (SELECT trace_id FROM ranked)
+GROUP BY trace_id
+ORDER BY last_activity_time DESC, trace_id DESC
+SETTINGS distributed_product_mode='allow', max_memory_usage=10000000000
+`, stmt)
+}
+
+// Span-only filter (no aggregate filter). WHERE is widened; HAVING has the gate + span
+// countIf pair but no trailing aggregate. `has_error = true` resolves to a
+// materialized-column predicate (not a map access). matched selects only the default order key.
+func TestBuild_FullSQL_TraceList_SpanFilterOnly(t *testing.T) {
+	b := newTestBuilder(t)
+	stmt, err := b.Build(context.Background(), valuer.UUID{}, testStartMs, testEndMs, qbtypes.RequestTypeTrace,
+		qbtypes.QueryBuilderQuery[qbtypes.TraceAggregation]{
+			Signal: telemetrytypes.SignalTraces, Source: telemetrytypes.SourceAI,
+			Filter: &qbtypes.Filter{Expression: "has_error = true"},
+			Limit:  20,
+		}, nil)
+	require.NoError(t, err)
+
+	requireSQLEqual(t, `
+WITH matched AS (
+    SELECT trace_id,
+        maxIf(timestamp, (mapContains(attributes_string, 'gen_ai.request.model') = true OR mapContains(attributes_string, 'gen_ai.tool.name') = true OR mapContains(attributes_string, 'gen_ai.agent.name') = true)) AS last_activity_time
+    FROM signoz_traces.distributed_signoz_index_v3
+    WHERE timestamp >= '1747947419000000000'
+      AND timestamp < '1747983448000000000'
+      AND ts_bucket_start >= 1747945619
+      AND ts_bucket_start <= 1747983448
+      AND ((mapContains(attributes_string, 'gen_ai.request.model') = true OR mapContains(attributes_string, 'gen_ai.tool.name') = true OR mapContains(attributes_string, 'gen_ai.agent.name') = true)
+        OR has_error = true)
+    GROUP BY trace_id
+    HAVING countIf((mapContains(attributes_string, 'gen_ai.request.model') = true OR mapContains(attributes_string, 'gen_ai.tool.name') = true OR mapContains(attributes_string, 'gen_ai.agent.name') = true)) > 0
+        AND countIf(has_error = true) > 0
+    ORDER BY last_activity_time DESC, trace_id DESC
+    LIMIT 20
+),
+ranked AS (
+    SELECT trace_id, min(start) AS t_start, max(end) AS t_end
+    FROM signoz_traces.distributed_trace_summary
+    WHERE trace_id GLOBAL IN (SELECT trace_id FROM matched)
+      AND end >= fromUnixTimestamp64Nano(1747947419000000000)
+      AND start < fromUnixTimestamp64Nano(1747983448000000000)
+    GROUP BY trace_id
+),
+buckets AS (
+    SELECT DISTINCT b AS ts_bucket
+    FROM ranked
+    ARRAY JOIN range(toUInt64(intDiv(toUnixTimestamp(t_start), 1800) * 1800 - 1800), toUInt64(intDiv(toUnixTimestamp(t_end), 1800) * 1800 + 1800), 1800) AS b
+)
+SELECT trace_id,
+    min(timestamp) AS start_time,
+    max(timestamp) AS end_time,
+    (max(toUnixTimestamp64Nano(timestamp) + duration_nano) - min(toUnixTimestamp64Nano(timestamp))) AS trace_duration_nano,
+    count() AS span_count,
+    anyIf(name, parent_span_id = '') AS root_span_name,
+    any(multiIf(multiIf(resource.service.name IS NOT NULL, resource.service.name::String, mapContains(resources_string, 'service.name'), resources_string['service.name'], NULL) IS NOT NULL, multiIf(resource.service.name IS NOT NULL, resource.service.name::String, mapContains(resources_string, 'service.name'), resources_string['service.name'], NULL), NULL)) AS service.name,
+    countIf(mapContains(attributes_string, 'gen_ai.request.model') = true) AS llm_call_count,
+    countIf(mapContains(attributes_string, 'gen_ai.tool.name') = true) AS tool_call_count,
+    uniqIf(multiIf(mapContains(attributes_string, 'gen_ai.tool.name') = true, attributes_string['gen_ai.tool.name'], NULL), mapContains(attributes_string, 'gen_ai.tool.name') = true) AS distinct_tool_count,
+    sum(multiIf(mapContains(attributes_number, 'gen_ai.usage.input_tokens') = true, toFloat64(attributes_number['gen_ai.usage.input_tokens']), NULL)) AS input_tokens,
+    sum(multiIf(mapContains(attributes_number, 'gen_ai.usage.output_tokens') = true, toFloat64(attributes_number['gen_ai.usage.output_tokens']), NULL)) AS output_tokens,
+    coalesce(sum(multiIf(mapContains(attributes_number, 'gen_ai.usage.input_tokens') = true, toFloat64(attributes_number['gen_ai.usage.input_tokens']), NULL)), 0) + coalesce(sum(multiIf(mapContains(attributes_number, 'gen_ai.usage.output_tokens') = true, toFloat64(attributes_number['gen_ai.usage.output_tokens']), NULL)), 0) AS total_tokens,
+    sum(multiIf(mapContains(attributes_number, '_signoz.gen_ai.total_cost') = true, toFloat64(attributes_number['_signoz.gen_ai.total_cost']), NULL)) AS estimated_cost_usd,
+    maxIf(duration_nano, mapContains(attributes_string, 'gen_ai.request.model') = true) AS max_llm_latency_ns,
+    countIf(has_error = true) AS error_count,
+    maxIf(timestamp, (mapContains(attributes_string, 'gen_ai.request.model') = true OR mapContains(attributes_string, 'gen_ai.tool.name') = true OR mapContains(attributes_string, 'gen_ai.agent.name') = true)) AS last_activity_time,
+    argMinIf(multiIf(mapContains(attributes_string, 'gen_ai.input.messages') = true, attributes_string['gen_ai.input.messages'], NULL), timestamp, mapContains(attributes_string, 'gen_ai.input.messages') = true) AS input,
+    argMaxIf(multiIf(mapContains(attributes_string, 'gen_ai.output.messages') = true, attributes_string['gen_ai.output.messages'], NULL), timestamp, mapContains(attributes_string, 'gen_ai.output.messages') = true) AS output
+FROM signoz_traces.distributed_signoz_index_v3
+WHERE ts_bucket_start GLOBAL IN (SELECT ts_bucket FROM buckets)
+  AND trace_id GLOBAL IN (SELECT trace_id FROM ranked)
+GROUP BY trace_id
+ORDER BY last_activity_time DESC, trace_id DESC
+SETTINGS distributed_product_mode='allow', max_memory_usage=10000000000
+`, stmt)
+}
+
+// Resource filter: a resource attribute in the filter is pulled into a __resource_filter
+// CTE (fingerprints matching the resource condition), and the `matched` scan is narrowed
+// by `resource_fingerprint GLOBAL IN (…)`. The resource key is dropped from the span
+// predicate (skipResourceFilter), so here there is no span-level existence check — the
+// prune stays the gate mask and the whole match is scoped to the resource fingerprints.
+func TestBuild_FullSQL_TraceList_ResourceFilter(t *testing.T) {
+	b := newTestBuilder(t)
+	stmt, err := b.Build(context.Background(), valuer.UUID{}, testStartMs, testEndMs, qbtypes.RequestTypeTrace,
+		qbtypes.QueryBuilderQuery[qbtypes.TraceAggregation]{
+			Signal: telemetrytypes.SignalTraces, Source: telemetrytypes.SourceAI,
+			Filter: &qbtypes.Filter{Expression: "resource.service.name = 'checkout'"},
+			Limit:  20,
+		}, nil)
+	require.NoError(t, err)
+
+	requireSQLEqual(t, `
+WITH __resource_filter AS (
+    SELECT fingerprint
+    FROM signoz_traces.distributed_traces_v3_resource
+    WHERE (simpleJSONExtractString(labels, 'service.name') = 'checkout' AND labels LIKE '%service.name%' AND labels LIKE '%service.name":"checkout%')
+      AND seen_at_ts_bucket_start >= 1747945619
+      AND seen_at_ts_bucket_start <= 1747983448
+    GROUP BY fingerprint
+),
+matched AS (
+    SELECT trace_id,
+        maxIf(timestamp, (mapContains(attributes_string, 'gen_ai.request.model') = true OR mapContains(attributes_string, 'gen_ai.tool.name') = true OR mapContains(attributes_string, 'gen_ai.agent.name') = true)) AS last_activity_time
+    FROM signoz_traces.distributed_signoz_index_v3
+    WHERE timestamp >= '1747947419000000000'
+      AND timestamp < '1747983448000000000'
+      AND ts_bucket_start >= 1747945619
+      AND ts_bucket_start <= 1747983448
+      AND ((mapContains(attributes_string, 'gen_ai.request.model') = true OR mapContains(attributes_string, 'gen_ai.tool.name') = true OR mapContains(attributes_string, 'gen_ai.agent.name') = true))
+      AND resource_fingerprint GLOBAL IN (SELECT fingerprint FROM __resource_filter)
+    GROUP BY trace_id
+    ORDER BY last_activity_time DESC, trace_id DESC
+    LIMIT 20
+),
+ranked AS (
+    SELECT trace_id, min(start) AS t_start, max(end) AS t_end
+    FROM signoz_traces.distributed_trace_summary
+    WHERE trace_id GLOBAL IN (SELECT trace_id FROM matched)
+      AND end >= fromUnixTimestamp64Nano(1747947419000000000)
+      AND start < fromUnixTimestamp64Nano(1747983448000000000)
+    GROUP BY trace_id
+),
+buckets AS (
+    SELECT DISTINCT b AS ts_bucket
+    FROM ranked
+    ARRAY JOIN range(toUInt64(intDiv(toUnixTimestamp(t_start), 1800) * 1800 - 1800), toUInt64(intDiv(toUnixTimestamp(t_end), 1800) * 1800 + 1800), 1800) AS b
+)
+SELECT trace_id,
+    min(timestamp) AS start_time,
+    max(timestamp) AS end_time,
+    (max(toUnixTimestamp64Nano(timestamp) + duration_nano) - min(toUnixTimestamp64Nano(timestamp))) AS trace_duration_nano,
+    count() AS span_count,
+    anyIf(name, parent_span_id = '') AS root_span_name,
+    any(multiIf(multiIf(resource.service.name IS NOT NULL, resource.service.name::String, mapContains(resources_string, 'service.name'), resources_string['service.name'], NULL) IS NOT NULL, multiIf(resource.service.name IS NOT NULL, resource.service.name::String, mapContains(resources_string, 'service.name'), resources_string['service.name'], NULL), NULL)) AS service.name,
+    countIf(mapContains(attributes_string, 'gen_ai.request.model') = true) AS llm_call_count,
+    countIf(mapContains(attributes_string, 'gen_ai.tool.name') = true) AS tool_call_count,
+    uniqIf(multiIf(mapContains(attributes_string, 'gen_ai.tool.name') = true, attributes_string['gen_ai.tool.name'], NULL), mapContains(attributes_string, 'gen_ai.tool.name') = true) AS distinct_tool_count,
+    sum(multiIf(mapContains(attributes_number, 'gen_ai.usage.input_tokens') = true, toFloat64(attributes_number['gen_ai.usage.input_tokens']), NULL)) AS input_tokens,
+    sum(multiIf(mapContains(attributes_number, 'gen_ai.usage.output_tokens') = true, toFloat64(attributes_number['gen_ai.usage.output_tokens']), NULL)) AS output_tokens,
+    coalesce(sum(multiIf(mapContains(attributes_number, 'gen_ai.usage.input_tokens') = true, toFloat64(attributes_number['gen_ai.usage.input_tokens']), NULL)), 0) + coalesce(sum(multiIf(mapContains(attributes_number, 'gen_ai.usage.output_tokens') = true, toFloat64(attributes_number['gen_ai.usage.output_tokens']), NULL)), 0) AS total_tokens,
+    sum(multiIf(mapContains(attributes_number, '_signoz.gen_ai.total_cost') = true, toFloat64(attributes_number['_signoz.gen_ai.total_cost']), NULL)) AS estimated_cost_usd,
+    maxIf(duration_nano, mapContains(attributes_string, 'gen_ai.request.model') = true) AS max_llm_latency_ns,
+    countIf(has_error = true) AS error_count,
+    maxIf(timestamp, (mapContains(attributes_string, 'gen_ai.request.model') = true OR mapContains(attributes_string, 'gen_ai.tool.name') = true OR mapContains(attributes_string, 'gen_ai.agent.name') = true)) AS last_activity_time,
+    argMinIf(multiIf(mapContains(attributes_string, 'gen_ai.input.messages') = true, attributes_string['gen_ai.input.messages'], NULL), timestamp, mapContains(attributes_string, 'gen_ai.input.messages') = true) AS input,
+    argMaxIf(multiIf(mapContains(attributes_string, 'gen_ai.output.messages') = true, attributes_string['gen_ai.output.messages'], NULL), timestamp, mapContains(attributes_string, 'gen_ai.output.messages') = true) AS output
+FROM signoz_traces.distributed_signoz_index_v3
+WHERE ts_bucket_start GLOBAL IN (SELECT ts_bucket FROM buckets)
+  AND trace_id GLOBAL IN (SELECT trace_id FROM ranked)
+GROUP BY trace_id
+ORDER BY last_activity_time DESC, trace_id DESC
+SETTINGS distributed_product_mode='allow', max_memory_usage=10000000000
+`, stmt)
+}
+
+// Mixed filter (two span predicates AND'd into one existence check + an aggregate) with
+// a two-key order on different aggregates than the filter. matched selects input_tokens
+// + last_activity_time (ORDER BY) and output_tokens (HAVING) — three of four; llm_call_count is not.
+func TestBuild_FullSQL_TraceList_MixedFiltersMultiOrder(t *testing.T) {
+	b := newTestBuilder(t)
+	stmt, err := b.Build(context.Background(), valuer.UUID{}, testStartMs, testEndMs, qbtypes.RequestTypeTrace,
+		qbtypes.QueryBuilderQuery[qbtypes.TraceAggregation]{
+			Signal: telemetrytypes.SignalTraces, Source: telemetrytypes.SourceAI,
+			Filter: &qbtypes.Filter{Expression: "gen_ai.request.model = 'gpt-4o' AND has_error = true AND output_tokens > 500"},
+			Order: []qbtypes.OrderBy{
+				{Key: qbtypes.OrderByKey{TelemetryFieldKey: telemetrytypes.TelemetryFieldKey{Name: "input_tokens"}}, Direction: qbtypes.OrderDirectionDesc},
+				{Key: qbtypes.OrderByKey{TelemetryFieldKey: telemetrytypes.TelemetryFieldKey{Name: "last_activity_time"}}, Direction: qbtypes.OrderDirectionAsc},
+			},
+			Limit: 15,
+		}, nil)
+	require.NoError(t, err)
+
+	requireSQLEqual(t, `
+WITH matched AS (
+    SELECT trace_id,
+        sum(multiIf(mapContains(attributes_number, 'gen_ai.usage.input_tokens') = true, toFloat64(attributes_number['gen_ai.usage.input_tokens']), NULL)) AS input_tokens,
+        sum(multiIf(mapContains(attributes_number, 'gen_ai.usage.output_tokens') = true, toFloat64(attributes_number['gen_ai.usage.output_tokens']), NULL)) AS output_tokens,
+        maxIf(timestamp, (mapContains(attributes_string, 'gen_ai.request.model') = true OR mapContains(attributes_string, 'gen_ai.tool.name') = true OR mapContains(attributes_string, 'gen_ai.agent.name') = true)) AS last_activity_time
+    FROM signoz_traces.distributed_signoz_index_v3
+    WHERE timestamp >= '1747947419000000000'
+      AND timestamp < '1747983448000000000'
+      AND ts_bucket_start >= 1747945619
+      AND ts_bucket_start <= 1747983448
+      AND ((mapContains(attributes_string, 'gen_ai.request.model') = true OR mapContains(attributes_string, 'gen_ai.tool.name') = true OR mapContains(attributes_string, 'gen_ai.agent.name') = true)
+        OR ((attributes_string['gen_ai.request.model'] = 'gpt-4o' AND mapContains(attributes_string, 'gen_ai.request.model') = true) AND has_error = true))
+    GROUP BY trace_id
+    HAVING countIf((mapContains(attributes_string, 'gen_ai.request.model') = true OR mapContains(attributes_string, 'gen_ai.tool.name') = true OR mapContains(attributes_string, 'gen_ai.agent.name') = true)) > 0
+        AND countIf(((attributes_string['gen_ai.request.model'] = 'gpt-4o' AND mapContains(attributes_string, 'gen_ai.request.model') = true) AND has_error = true)) > 0
+        AND output_tokens > 500
+    ORDER BY input_tokens DESC, last_activity_time ASC, trace_id DESC
+    LIMIT 15
+),
+ranked AS (
+    SELECT trace_id, min(start) AS t_start, max(end) AS t_end
+    FROM signoz_traces.distributed_trace_summary
+    WHERE trace_id GLOBAL IN (SELECT trace_id FROM matched)
+      AND end >= fromUnixTimestamp64Nano(1747947419000000000)
+      AND start < fromUnixTimestamp64Nano(1747983448000000000)
+    GROUP BY trace_id
+),
+buckets AS (
+    SELECT DISTINCT b AS ts_bucket
+    FROM ranked
+    ARRAY JOIN range(toUInt64(intDiv(toUnixTimestamp(t_start), 1800) * 1800 - 1800), toUInt64(intDiv(toUnixTimestamp(t_end), 1800) * 1800 + 1800), 1800) AS b
+)
+SELECT trace_id,
+    min(timestamp) AS start_time,
+    max(timestamp) AS end_time,
+    (max(toUnixTimestamp64Nano(timestamp) + duration_nano) - min(toUnixTimestamp64Nano(timestamp))) AS trace_duration_nano,
+    count() AS span_count,
+    anyIf(name, parent_span_id = '') AS root_span_name,
+    any(multiIf(multiIf(resource.service.name IS NOT NULL, resource.service.name::String, mapContains(resources_string, 'service.name'), resources_string['service.name'], NULL) IS NOT NULL, multiIf(resource.service.name IS NOT NULL, resource.service.name::String, mapContains(resources_string, 'service.name'), resources_string['service.name'], NULL), NULL)) AS service.name,
+    countIf(mapContains(attributes_string, 'gen_ai.request.model') = true) AS llm_call_count,
+    countIf(mapContains(attributes_string, 'gen_ai.tool.name') = true) AS tool_call_count,
+    uniqIf(multiIf(mapContains(attributes_string, 'gen_ai.tool.name') = true, attributes_string['gen_ai.tool.name'], NULL), mapContains(attributes_string, 'gen_ai.tool.name') = true) AS distinct_tool_count,
+    sum(multiIf(mapContains(attributes_number, 'gen_ai.usage.input_tokens') = true, toFloat64(attributes_number['gen_ai.usage.input_tokens']), NULL)) AS input_tokens,
+    sum(multiIf(mapContains(attributes_number, 'gen_ai.usage.output_tokens') = true, toFloat64(attributes_number['gen_ai.usage.output_tokens']), NULL)) AS output_tokens,
+    coalesce(sum(multiIf(mapContains(attributes_number, 'gen_ai.usage.input_tokens') = true, toFloat64(attributes_number['gen_ai.usage.input_tokens']), NULL)), 0) + coalesce(sum(multiIf(mapContains(attributes_number, 'gen_ai.usage.output_tokens') = true, toFloat64(attributes_number['gen_ai.usage.output_tokens']), NULL)), 0) AS total_tokens,
+    sum(multiIf(mapContains(attributes_number, '_signoz.gen_ai.total_cost') = true, toFloat64(attributes_number['_signoz.gen_ai.total_cost']), NULL)) AS estimated_cost_usd,
+    maxIf(duration_nano, mapContains(attributes_string, 'gen_ai.request.model') = true) AS max_llm_latency_ns,
+    countIf(has_error = true) AS error_count,
+    maxIf(timestamp, (mapContains(attributes_string, 'gen_ai.request.model') = true OR mapContains(attributes_string, 'gen_ai.tool.name') = true OR mapContains(attributes_string, 'gen_ai.agent.name') = true)) AS last_activity_time,
+    argMinIf(multiIf(mapContains(attributes_string, 'gen_ai.input.messages') = true, attributes_string['gen_ai.input.messages'], NULL), timestamp, mapContains(attributes_string, 'gen_ai.input.messages') = true) AS input,
+    argMaxIf(multiIf(mapContains(attributes_string, 'gen_ai.output.messages') = true, attributes_string['gen_ai.output.messages'], NULL), timestamp, mapContains(attributes_string, 'gen_ai.output.messages') = true) AS output
+FROM signoz_traces.distributed_signoz_index_v3
+WHERE ts_bucket_start GLOBAL IN (SELECT ts_bucket FROM buckets)
+  AND trace_id GLOBAL IN (SELECT trace_id FROM ranked)
+GROUP BY trace_id
+ORDER BY input_tokens DESC, last_activity_time ASC, trace_id DESC
+SETTINGS distributed_product_mode='allow', max_memory_usage=10000000000
+`, stmt)
+}
+
+// Span list (requestType raw): delegated to the traces builder with the gate ANDed
+// into the user filter, so only gen_ai spans matching the filter come back. Standard
+// span columns, single SELECT (no CTE pipeline).
+func TestBuild_FullSQL_SpanList_Raw(t *testing.T) {
+	b := newTestBuilder(t)
+	stmt, err := b.Build(context.Background(), valuer.UUID{}, testStartMs, testEndMs, qbtypes.RequestTypeRaw,
+		qbtypes.QueryBuilderQuery[qbtypes.TraceAggregation]{
+			Signal: telemetrytypes.SignalTraces, Source: telemetrytypes.SourceAI,
+			Filter: &qbtypes.Filter{Expression: "gen_ai.request.model = 'gpt-4o-mini'"},
+			Limit:  10,
+		}, nil)
+	require.NoError(t, err)
+
+	requireSQLEqual(t, `
+SELECT timestamp AS __SELECT_KEY_0_timestamp, trace_id AS __SELECT_KEY_1_trace_id, span_id AS __SELECT_KEY_2_span_id,
+    trace_state AS __SELECT_KEY_3_trace_state, parent_span_id AS __SELECT_KEY_4_parent_span_id, flags AS __SELECT_KEY_5_flags,
+    name AS __SELECT_KEY_6_name, kind AS __SELECT_KEY_7_kind, kind_string AS __SELECT_KEY_8_kind_string, duration_nano AS __SELECT_KEY_9_duration_nano,
+    status_code AS __SELECT_KEY_10_status_code, status_message AS __SELECT_KEY_11_status_message,
+    status_code_string AS __SELECT_KEY_12_status_code_string, events AS __SELECT_KEY_13_events, links AS __SELECT_KEY_14_links,
+    response_status_code AS __SELECT_KEY_15_response_status_code, external_http_url AS __SELECT_KEY_16_external_http_url,
+    http_url AS __SELECT_KEY_17_http_url, external_http_method AS __SELECT_KEY_18_external_http_method,
+    http_method AS __SELECT_KEY_19_http_method, http_host AS __SELECT_KEY_20_http_host, db_name AS __SELECT_KEY_21_db_name,
+    db_operation AS __SELECT_KEY_22_db_operation, has_error AS __SELECT_KEY_23_has_error, is_remote AS __SELECT_KEY_24_is_remote,
+    attributes_string, attributes_number, attributes_bool, resources_string
+FROM signoz_traces.distributed_signoz_index_v3
+WHERE (((mapContains(attributes_string, 'gen_ai.request.model') = true
+        OR mapContains(attributes_string, 'gen_ai.tool.name') = true
+        OR mapContains(attributes_string, 'gen_ai.agent.name') = true))
+    AND ((attributes_string['gen_ai.request.model'] = 'gpt-4o-mini'
+        AND mapContains(attributes_string, 'gen_ai.request.model') = true)))
+  AND timestamp >= '1747947419000000000'
+  AND timestamp < '1747983448000000000'
+  AND ts_bucket_start >= 1747945619
+  AND ts_bucket_start <= 1747983448
+LIMIT 10
+`, stmt)
+}
+
+// ---------------------------------------------------------------------------
+// Behavior / branch tests not covered by the goldens above
+// ---------------------------------------------------------------------------
+
+// A filter mixing a resource attribute with a span-level and an aggregate condition:
+// the resource key routes into __resource_filter (fingerprint prune), the span key stays
+// as a countIf existence check, and the aggregate becomes a HAVING — all AND-combined.
+// service.name (resource context) comes from otelKeysMap.
+func TestBuild_TraceList_ResourcePlusSpanPlusAggregateFilter(t *testing.T) {
+	b := newTestBuilder(t)
+	stmt, err := b.Build(context.Background(), valuer.UUID{}, testStartMs, testEndMs, qbtypes.RequestTypeTrace,
+		qbtypes.QueryBuilderQuery[qbtypes.TraceAggregation]{
+			Signal: telemetrytypes.SignalTraces, Source: telemetrytypes.SourceAI,
+			Filter: &qbtypes.Filter{Expression: "resource.service.name = 'checkout' AND has_error = true AND output_tokens > 1000"},
+			Limit:  10,
+		}, nil)
+	require.NoError(t, err)
+
+	got := renderSQL(t, stmt)
+	// resource condition -> fingerprint CTE + prune, not filtered on the span index
+	// (the service.name output column still reads the resource map, hence the = form).
+	require.Contains(t, got, "__resource_filter AS (")
+	require.Contains(t, got, "resource_fingerprint GLOBAL IN (SELECT fingerprint FROM __resource_filter)")
+	require.NotContains(t, got, "resources_string['service.name'] = 'checkout'")
+	// span condition -> existence check in matched HAVING.
+	require.Contains(t, got, "countIf(has_error = true) > 0")
+	// aggregate condition -> HAVING on the matched aggregate alias.
+	require.Contains(t, got, "output_tokens")
+}
+
+// The resolver-unset (nil) fallback is covered in pkg/telemetryscopedtraces, which
+// can construct that builder state directly.
+
+// Trace-level and span-level predicates may not be OR-combined.
+func TestBuild_TraceList_TraceOrSpanMixRejected(t *testing.T) {
+	b := newTestBuilder(t)
+	query := qbtypes.QueryBuilderQuery[qbtypes.TraceAggregation]{
+		Signal: telemetrytypes.SignalTraces,
+		Source: telemetrytypes.SourceAI,
+		Filter: &qbtypes.Filter{Expression: "trace.output_tokens > 1000 OR gen_ai.request.model = 'x'"},
+		Limit:  10,
+	}
+	_, err := b.Build(context.Background(), valuer.UUID{}, testStartMs, testEndMs, qbtypes.RequestTypeTrace, query, nil)
+	require.Error(t, err)
+	require.Contains(t, err.Error(), "cannot be combined")
+}
+
+// An output-only aggregate (span_count / trace_duration_nano) can be displayed but not
+// used in the aggregate filter or ORDER BY — it is not computable in the matched pass.
+func TestBuild_TraceList_OutputOnlyAggregateRejected(t *testing.T) {
+	b := newTestBuilder(t)
+
+	// filter by span_count -> rejected
+	_, err := b.Build(context.Background(), valuer.UUID{}, testStartMs, testEndMs, qbtypes.RequestTypeTrace,
+		qbtypes.QueryBuilderQuery[qbtypes.TraceAggregation]{
+			Signal: telemetrytypes.SignalTraces, Source: telemetrytypes.SourceAI,
+			Filter: &qbtypes.Filter{Expression: "span_count > 3"},
+		}, nil)
+	require.Error(t, err)
+	require.Contains(t, err.Error(), "span_count")
+
+	// order by trace_duration_nano -> rejected
+	_, err = b.Build(context.Background(), valuer.UUID{}, testStartMs, testEndMs, qbtypes.RequestTypeTrace,
+		qbtypes.QueryBuilderQuery[qbtypes.TraceAggregation]{
+			Signal: telemetrytypes.SignalTraces, Source: telemetrytypes.SourceAI,
+			Order: []qbtypes.OrderBy{{Key: qbtypes.OrderByKey{TelemetryFieldKey: telemetrytypes.TelemetryFieldKey{Name: "trace_duration_nano"}}, Direction: qbtypes.OrderDirectionDesc}},
+		}, nil)
+	require.Error(t, err)
+	require.Contains(t, err.Error(), "unsupported order key")
+}
+
+// duration_nano no longer names an aggregate (the trace column is trace_duration_nano),
+// so a bare filter on it is span-level like everywhere else in the product: the trace
+// matches when any span exceeds the duration.
+func TestBuild_TraceList_SpanDurationFilterIsSpanLevel(t *testing.T) {
+	keys := otelKeysMap()
+	keys["duration_nano"] = []*telemetrytypes.TelemetryFieldKey{{
+		Name:          "duration_nano",
+		Signal:        telemetrytypes.SignalTraces,
+		FieldContext:  telemetrytypes.FieldContextSpan,
+		FieldDataType: telemetrytypes.FieldDataTypeNumber,
+	}}
+	b := newTestBuilderWithKeys(t, keys)
+	stmt, err := b.Build(context.Background(), valuer.UUID{}, testStartMs, testEndMs, qbtypes.RequestTypeTrace,
+		qbtypes.QueryBuilderQuery[qbtypes.TraceAggregation]{
+			Signal: telemetrytypes.SignalTraces, Source: telemetrytypes.SourceAI,
+			Filter: &qbtypes.Filter{Expression: "duration_nano > 1000000"},
+			Limit:  10,
+		}, nil)
+	require.NoError(t, err)
+
+	got := renderSQL(t, stmt)
+	require.Contains(t, got, "countIf(duration_nano > 1e+06) > 0")
+	require.NotContains(t, got, "HAVING trace_duration_nano")
+}
+
+// A HAVING referencing a non-aggregate column is rejected.
+func TestBuild_TraceList_Having_UnknownColumn(t *testing.T) {
+	b := newTestBuilder(t)
+	query := qbtypes.QueryBuilderQuery[qbtypes.TraceAggregation]{
+		Signal: telemetrytypes.SignalTraces,
+		Source: telemetrytypes.SourceAI,
+		Having: &qbtypes.Having{Expression: "service.name > 1"}, // not an aggregate column
+		Limit:  10,
+	}
+	_, err := b.Build(context.Background(), valuer.UUID{}, testStartMs, testEndMs, qbtypes.RequestTypeTrace, query, nil)
+	require.Error(t, err)
+}
+
+// Ordering by an unknown key is rejected.
+func TestBuild_TraceList_UnsupportedOrderKey(t *testing.T) {
+	b := newTestBuilder(t)
+	query := qbtypes.QueryBuilderQuery[qbtypes.TraceAggregation]{
+		Signal: telemetrytypes.SignalTraces,
+		Source: telemetrytypes.SourceAI,
+		Order: []qbtypes.OrderBy{
+			{Key: qbtypes.OrderByKey{TelemetryFieldKey: telemetrytypes.TelemetryFieldKey{Name: "http.request.method"}}, Direction: qbtypes.OrderDirectionDesc},
+		},
+	}
+	_, err := b.Build(context.Background(), valuer.UUID{}, testStartMs, testEndMs, qbtypes.RequestTypeTrace, query, nil)
+	require.Error(t, err)
+	require.Contains(t, err.Error(), "unsupported order key")
+}
+
+// With no limit set, the builder applies the default of 100.
+func TestBuild_TraceList_DefaultLimit(t *testing.T) {
+	b := newTestBuilder(t)
+	query := qbtypes.QueryBuilderQuery[qbtypes.TraceAggregation]{
+		Signal: telemetrytypes.SignalTraces,
+		Source: telemetrytypes.SourceAI,
+	}
+	stmt, err := b.Build(context.Background(), valuer.UUID{}, testStartMs, testEndMs, qbtypes.RequestTypeTrace, query, nil)
+	require.NoError(t, err)
+	require.Contains(t, stmt.Query, "LIMIT ?")
+	require.Contains(t, stmt.Args, 100)
+}
+
+// Only trace list and span list (raw) are supported; distribution is not.
+func TestBuild_UnsupportedRequestType(t *testing.T) {
+	b := newTestBuilder(t)
+	query := qbtypes.QueryBuilderQuery[qbtypes.TraceAggregation]{
+		Signal: telemetrytypes.SignalTraces,
+		Source: telemetrytypes.SourceAI,
+		Aggregations: []qbtypes.TraceAggregation{
+			{Expression: "count()"},
+		},
+	}
+	_, err := b.Build(context.Background(), valuer.UUID{}, testStartMs, testEndMs, qbtypes.RequestTypeDistribution, query, nil)
+	require.ErrorIs(t, err, scopedtraces.ErrUnsupportedRequestType)
+}
+
+// A gate key ingested under several data types (e.g. string + number from a
+// misbehaving SDK) contributes ALL variants to the mask, OR-combined — not just
+// the first — matching the standard visitor's EXISTS handling.
+func TestBuild_TraceList_MultiVariantGateKey(t *testing.T) {
+	keys := otelKeysMap()
+	keys[telemetrytypes.GenAIToolName] = append(keys[telemetrytypes.GenAIToolName], &telemetrytypes.TelemetryFieldKey{
+		Name:          telemetrytypes.GenAIToolName,
+		Signal:        telemetrytypes.SignalTraces,
+		FieldContext:  telemetrytypes.FieldContextAttribute,
+		FieldDataType: telemetrytypes.FieldDataTypeFloat64,
+	})
+	b := newTestBuilderWithKeys(t, keys)
+	stmt, err := b.Build(context.Background(), valuer.UUID{}, testStartMs, testEndMs, qbtypes.RequestTypeTrace,
+		qbtypes.QueryBuilderQuery[qbtypes.TraceAggregation]{
+			Signal: telemetrytypes.SignalTraces, Source: telemetrytypes.SourceAI, Limit: 10,
+		}, nil)
+	require.NoError(t, err)
+
+	got := renderSQL(t, stmt)
+	require.Contains(t, got, "mapContains(attributes_string, 'gen_ai.tool.name') = true OR mapContains(attributes_number, 'gen_ai.tool.name') = true")
+}
+
+// `trace.` parses as the trace field context and marks a trace-level aggregate; the
+// legacy `tracefield.` spelling is explicitly rejected (filter and having alike), and
+// an output-only aggregate under the context gets the targeted rejection rather than
+// an unknown-span-field failure.
+func TestBuild_TraceList_TraceContextPrefix(t *testing.T) {
+	b := newTestBuilder(t)
+	build := func(q qbtypes.QueryBuilderQuery[qbtypes.TraceAggregation]) (*qbtypes.Statement, error) {
+		q.Signal, q.Source, q.Limit = telemetrytypes.SignalTraces, telemetrytypes.SourceAI, 20
+		return b.Build(context.Background(), valuer.UUID{}, testStartMs, testEndMs, qbtypes.RequestTypeTrace, q, nil)
+	}
+
+	_, err := build(qbtypes.QueryBuilderQuery[qbtypes.TraceAggregation]{
+		Filter: &qbtypes.Filter{Expression: "trace.output_tokens > 1000"}})
+	require.NoError(t, err)
+
+	_, err = build(qbtypes.QueryBuilderQuery[qbtypes.TraceAggregation]{
+		Filter: &qbtypes.Filter{Expression: "tracefield.output_tokens > 1000"}})
+	require.Error(t, err)
+	require.Contains(t, err.Error(), `use the "trace." prefix`)
+
+	_, err = build(qbtypes.QueryBuilderQuery[qbtypes.TraceAggregation]{
+		Having: &qbtypes.Having{Expression: "tracefield.output_tokens > 1000"}})
+	require.Error(t, err)
+	require.Contains(t, err.Error(), `use the "trace." prefix`)
+
+	_, err = build(qbtypes.QueryBuilderQuery[qbtypes.TraceAggregation]{
+		Filter: &qbtypes.Filter{Expression: "trace.span_count > 3"}})
+	require.Error(t, err)
+	require.Contains(t, err.Error(), "cannot be used")
+}
+
+// Query variables in a trace-level condition are substituted into the HAVING (the
+// span path binds them via PrepareWhereClause; the HAVING is a text rewrite).
+func TestBuild_TraceList_VariableInAggregateFilter(t *testing.T) {
+	b := newTestBuilder(t)
+	build := func(expr string, vars map[string]qbtypes.VariableItem) (*qbtypes.Statement, error) {
+		return b.Build(context.Background(), valuer.UUID{}, testStartMs, testEndMs, qbtypes.RequestTypeTrace,
+			qbtypes.QueryBuilderQuery[qbtypes.TraceAggregation]{
+				Signal: telemetrytypes.SignalTraces, Source: telemetrytypes.SourceAI,
+				Filter: &qbtypes.Filter{Expression: expr},
+				Limit:  20,
+			}, vars)
+	}
+
+	// scalar variable -> literal in HAVING
+	stmt, err := build("trace.output_tokens > $threshold",
+		map[string]qbtypes.VariableItem{"threshold": {Value: 700}})
+	require.NoError(t, err)
+	require.Contains(t, stmt.Query, "HAVING output_tokens > 700")
+
+	// list variable with IN
+	stmt, err = build("trace.llm_call_count IN $counts",
+		map[string]qbtypes.VariableItem{"counts": {Value: []any{1, 2}}})
+	require.NoError(t, err)
+	require.Contains(t, stmt.Query, "HAVING llm_call_count IN")
+
+	// dynamic __all__ -> condition dropped, no HAVING at all
+	stmt, err = build("trace.output_tokens > $threshold",
+		map[string]qbtypes.VariableItem{"threshold": {Type: qbtypes.DynamicVariableType, Value: "__all__"}})
+	require.NoError(t, err)
+	require.NotContains(t, stmt.Query, "HAVING")
+
+	// unresolved variable -> rejected, not compared as a literal
+	_, err = build("trace.output_tokens > $missing", map[string]qbtypes.VariableItem{"other": {Value: 1}})
+	require.Error(t, err)
+}

pkg/telemetryaudit/condition_builder.go

@@ -9,6 +9,7 @@ import (
 	"github.com/SigNoz/signoz/pkg/querybuilder"
 	qbtypes "github.com/SigNoz/signoz/pkg/types/querybuildertypes/querybuildertypesv5"
 	"github.com/SigNoz/signoz/pkg/types/telemetrytypes"
+	"github.com/SigNoz/signoz/pkg/valuer"
 	"github.com/huandu/go-sqlbuilder"
 )
 
@@ -172,6 +173,7 @@ func (c *conditionBuilder) conditionFor(
 
 func (c *conditionBuilder) ConditionFor(
 	ctx context.Context,
+	orgID valuer.UUID,
 	startNs uint64,
 	endNs uint64,
 	key *telemetrytypes.TelemetryFieldKey,

pkg/telemetryaudit/field_mapper.go

@@ -8,6 +8,7 @@ import (
 	"github.com/SigNoz/signoz/pkg/errors"
 	qbtypes "github.com/SigNoz/signoz/pkg/types/querybuildertypes/querybuildertypesv5"
 	"github.com/SigNoz/signoz/pkg/types/telemetrytypes"
+	"github.com/SigNoz/signoz/pkg/valuer"
 	"github.com/huandu/go-sqlbuilder"
 
 	"golang.org/x/exp/maps"
@@ -97,6 +98,7 @@ func (m *fieldMapper) ColumnFor(ctx context.Context, _, _ uint64, key *telemetry
 
 func (m *fieldMapper) ColumnExpressionFor(
 	ctx context.Context,
+	_ valuer.UUID,
 	tsStart, tsEnd uint64,
 	field *telemetrytypes.TelemetryFieldKey,
 	keys map[string][]*telemetrytypes.TelemetryFieldKey,
@@ -119,3 +121,9 @@ func (m *fieldMapper) ColumnExpressionFor(
 
 	return fmt.Sprintf("%s AS `%s`", sqlbuilder.Escape(fieldExpression), field.Name), nil
 }
+
+// CandidateKeys returns nil: audit has no synthesize-on-unknown-key fallback, so an
+// unknown key stays unresolved and the caller errors.
+func (m *fieldMapper) CandidateKeys(_ context.Context, _ valuer.UUID, _ *telemetrytypes.TelemetryFieldKey, _ any, _ map[string][]*telemetrytypes.TelemetryFieldKey) []*telemetrytypes.TelemetryFieldKey {
+	return nil
+}

pkg/telemetryaudit/statement_builder.go

@@ -13,6 +13,7 @@ import (
 	"github.com/SigNoz/signoz/pkg/telemetryresourcefilter"
 	qbtypes "github.com/SigNoz/signoz/pkg/types/querybuildertypes/querybuildertypesv5"
 	"github.com/SigNoz/signoz/pkg/types/telemetrytypes"
+	"github.com/SigNoz/signoz/pkg/valuer"
 	"github.com/huandu/go-sqlbuilder"
 )
 
@@ -66,6 +67,7 @@ func NewAuditQueryStatementBuilder(
 
 func (b *auditQueryStatementBuilder) Build(
 	ctx context.Context,
+	orgID valuer.UUID,
 	start uint64,
 	end uint64,
 	requestType qbtypes.RequestType,
@@ -88,11 +90,11 @@ func (b *auditQueryStatementBuilder) Build(
 	var stmt *qbtypes.Statement
 	switch requestType {
 	case qbtypes.RequestTypeRaw, qbtypes.RequestTypeRawStream:
-		stmt, err = b.buildListQuery(ctx, q, query, start, end, keys, variables)
+		stmt, err = b.buildListQuery(ctx, orgID, q, query, start, end, keys, variables)
 	case qbtypes.RequestTypeTimeSeries:
-		stmt, err = b.buildTimeSeriesQuery(ctx, q, query, start, end, keys, variables)
+		stmt, err = b.buildTimeSeriesQuery(ctx, orgID, q, query, start, end, keys, variables)
 	case qbtypes.RequestTypeScalar:
-		stmt, err = b.buildScalarQuery(ctx, q, query, start, end, keys, false, variables)
+		stmt, err = b.buildScalarQuery(ctx, orgID, q, query, start, end, keys, false, variables)
 	default:
 		return nil, errors.NewInvalidInputf(errors.CodeInvalidInput, "unsupported request type: %s", requestType)
 	}
@@ -201,6 +203,7 @@ func (b *auditQueryStatementBuilder) adjustKey(key *telemetrytypes.TelemetryFiel
 
 func (b *auditQueryStatementBuilder) buildListQuery(
 	ctx context.Context,
+	orgID valuer.UUID,
 	sb *sqlbuilder.SelectBuilder,
 	query qbtypes.QueryBuilderQuery[qbtypes.LogAggregation],
 	start, end uint64,
@@ -212,7 +215,7 @@ func (b *auditQueryStatementBuilder) buildListQuery(
 		cteArgs      [][]any
 	)
 
-	if frag, args, err := b.maybeAttachResourceFilter(ctx, sb, query, start, end, variables); err != nil {
+	if frag, args, err := b.maybeAttachResourceFilter(ctx, orgID, sb, query, start, end, variables); err != nil {
 		return nil, err
 	} else if frag != "" {
 		cteFragments = append(cteFragments, frag)
@@ -242,7 +245,7 @@ func (b *auditQueryStatementBuilder) buildListQuery(
 				continue
 			}
 
-			colExpr, err := b.fm.ColumnExpressionFor(ctx, start, end, &query.SelectFields[index], keys)
+			colExpr, err := b.fm.ColumnExpressionFor(ctx, orgID, start, end, &query.SelectFields[index], keys)
 			if err != nil {
 				return nil, err
 			}
@@ -252,13 +255,13 @@ func (b *auditQueryStatementBuilder) buildListQuery(
 
 	sb.From(fmt.Sprintf("%s.%s", DBName, AuditLogsTableName))
 
-	preparedWhereClause, err := b.addFilterCondition(ctx, sb, start, end, query, keys, variables)
+	preparedWhereClause, err := b.addFilterCondition(ctx, orgID, sb, start, end, query, keys, variables)
 	if err != nil {
 		return nil, err
 	}
 
 	for _, orderBy := range query.Order {
-		colExpr, err := b.fm.ColumnExpressionFor(ctx, start, end, &orderBy.Key.TelemetryFieldKey, keys)
+		colExpr, err := b.fm.ColumnExpressionFor(ctx, orgID, start, end, &orderBy.Key.TelemetryFieldKey, keys)
 		if err != nil {
 			return nil, err
 		}
@@ -292,6 +295,7 @@ func (b *auditQueryStatementBuilder) buildListQuery(
 
 func (b *auditQueryStatementBuilder) buildTimeSeriesQuery(
 	ctx context.Context,
+	orgID valuer.UUID,
 	sb *sqlbuilder.SelectBuilder,
 	query qbtypes.QueryBuilderQuery[qbtypes.LogAggregation],
 	start, end uint64,
@@ -303,7 +307,7 @@ func (b *auditQueryStatementBuilder) buildTimeSeriesQuery(
 		cteArgs      [][]any
 	)
 
-	if frag, args, err := b.maybeAttachResourceFilter(ctx, sb, query, start, end, variables); err != nil {
+	if frag, args, err := b.maybeAttachResourceFilter(ctx, orgID, sb, query, start, end, variables); err != nil {
 		return nil, err
 	} else if frag != "" {
 		cteFragments = append(cteFragments, frag)
@@ -319,7 +323,7 @@ func (b *auditQueryStatementBuilder) buildTimeSeriesQuery(
 
 	fieldNames := make([]string, 0, len(query.GroupBy))
 	for _, gb := range query.GroupBy {
-		expr, args, err := querybuilder.CollisionHandledFinalExpr(ctx, start, end, &gb.TelemetryFieldKey, b.fm, b.cb, keys, telemetrytypes.FieldDataTypeString, b.jsonKeyToKey, false)
+		expr, args, err := querybuilder.CollisionHandledFinalExpr(ctx, orgID, start, end, &gb.TelemetryFieldKey, b.fm, b.cb, keys, telemetrytypes.FieldDataTypeString, b.jsonKeyToKey, false)
 		if err != nil {
 			return nil, err
 		}
@@ -332,7 +336,7 @@ func (b *auditQueryStatementBuilder) buildTimeSeriesQuery(
 
 	allAggChArgs := make([]any, 0)
 	for i, agg := range query.Aggregations {
-		rewritten, chArgs, err := b.aggExprRewriter.Rewrite(ctx, start, end, agg.Expression, uint64(query.StepInterval.Seconds()), keys)
+		rewritten, chArgs, err := b.aggExprRewriter.Rewrite(ctx, orgID, start, end, agg.Expression, uint64(query.StepInterval.Seconds()), keys)
 		if err != nil {
 			return nil, err
 		}
@@ -342,7 +346,7 @@ func (b *auditQueryStatementBuilder) buildTimeSeriesQuery(
 
 	sb.From(fmt.Sprintf("%s.%s", DBName, AuditLogsTableName))
 
-	preparedWhereClause, err := b.addFilterCondition(ctx, sb, start, end, query, keys, variables)
+	preparedWhereClause, err := b.addFilterCondition(ctx, orgID, sb, start, end, query, keys, variables)
 	if err != nil {
 		return nil, err
 	}
@@ -352,7 +356,7 @@ func (b *auditQueryStatementBuilder) buildTimeSeriesQuery(
 
 	if query.Limit > 0 && len(query.GroupBy) > 0 {
 		cteSB := sqlbuilder.NewSelectBuilder()
-		cteStmt, err := b.buildScalarQuery(ctx, cteSB, query, start, end, keys, true, variables)
+		cteStmt, err := b.buildScalarQuery(ctx, orgID, cteSB, query, start, end, keys, true, variables)
 		if err != nil {
 			return nil, err
 		}
@@ -430,6 +434,7 @@ func (b *auditQueryStatementBuilder) buildTimeSeriesQuery(
 
 func (b *auditQueryStatementBuilder) buildScalarQuery(
 	ctx context.Context,
+	orgID valuer.UUID,
 	sb *sqlbuilder.SelectBuilder,
 	query qbtypes.QueryBuilderQuery[qbtypes.LogAggregation],
 	start, end uint64,
@@ -442,7 +447,7 @@ func (b *auditQueryStatementBuilder) buildScalarQuery(
 		cteArgs      [][]any
 	)
 
-	if frag, args, err := b.maybeAttachResourceFilter(ctx, sb, query, start, end, variables); err != nil {
+	if frag, args, err := b.maybeAttachResourceFilter(ctx, orgID, sb, query, start, end, variables); err != nil {
 		return nil, err
 	} else if frag != "" && !skipResourceCTE {
 		cteFragments = append(cteFragments, frag)
@@ -454,7 +459,7 @@ func (b *auditQueryStatementBuilder) buildScalarQuery(
 	var allGroupByArgs []any
 
 	for _, gb := range query.GroupBy {
-		expr, args, err := querybuilder.CollisionHandledFinalExpr(ctx, start, end, &gb.TelemetryFieldKey, b.fm, b.cb, keys, telemetrytypes.FieldDataTypeString, b.jsonKeyToKey, false)
+		expr, args, err := querybuilder.CollisionHandledFinalExpr(ctx, orgID, start, end, &gb.TelemetryFieldKey, b.fm, b.cb, keys, telemetrytypes.FieldDataTypeString, b.jsonKeyToKey, false)
 		if err != nil {
 			return nil, err
 		}
@@ -469,7 +474,7 @@ func (b *auditQueryStatementBuilder) buildScalarQuery(
 	if len(query.Aggregations) > 0 {
 		for idx := range query.Aggregations {
 			aggExpr := query.Aggregations[idx]
-			rewritten, chArgs, err := b.aggExprRewriter.Rewrite(ctx, start, end, aggExpr.Expression, rateInterval, keys)
+			rewritten, chArgs, err := b.aggExprRewriter.Rewrite(ctx, orgID, start, end, aggExpr.Expression, rateInterval, keys)
 			if err != nil {
 				return nil, err
 			}
@@ -480,7 +485,7 @@ func (b *auditQueryStatementBuilder) buildScalarQuery(
 
 	sb.From(fmt.Sprintf("%s.%s", DBName, AuditLogsTableName))
 
-	preparedWhereClause, err := b.addFilterCondition(ctx, sb, start, end, query, keys, variables)
+	preparedWhereClause, err := b.addFilterCondition(ctx, orgID, sb, start, end, query, keys, variables)
 	if err != nil {
 		return nil, err
 	}
@@ -532,6 +537,7 @@ func (b *auditQueryStatementBuilder) buildScalarQuery(
 
 func (b *auditQueryStatementBuilder) addFilterCondition(
 	ctx context.Context,
+	orgID valuer.UUID,
 	sb *sqlbuilder.SelectBuilder,
 	start, end uint64,
 	query qbtypes.QueryBuilderQuery[qbtypes.LogAggregation],
@@ -544,6 +550,7 @@ func (b *auditQueryStatementBuilder) addFilterCondition(
 	if query.Filter != nil && query.Filter.Expression != "" {
 		preparedWhereClause, err = querybuilder.PrepareWhereClause(query.Filter.Expression, querybuilder.FilterExprVisitorOpts{
 			Context:            ctx,
+			OrgID:              orgID,
 			Logger:             b.logger,
 			FieldMapper:        b.fm,
 			ConditionBuilder:   b.cb,
@@ -591,12 +598,13 @@ func aggOrderBy(k qbtypes.OrderBy, q qbtypes.QueryBuilderQuery[qbtypes.LogAggreg
 
 func (b *auditQueryStatementBuilder) maybeAttachResourceFilter(
 	ctx context.Context,
+	orgID valuer.UUID,
 	sb *sqlbuilder.SelectBuilder,
 	query qbtypes.QueryBuilderQuery[qbtypes.LogAggregation],
 	start, end uint64,
 	variables map[string]qbtypes.VariableItem,
 ) (cteSQL string, cteArgs []any, err error) {
-	stmt, err := b.resourceFilterStmtBuilder.Build(ctx, start, end, qbtypes.RequestTypeRaw, query, variables)
+	stmt, err := b.resourceFilterStmtBuilder.Build(ctx, orgID, start, end, qbtypes.RequestTypeRaw, query, variables)
 	if err != nil {
 		return "", nil, err
 	}

pkg/telemetryaudit/statement_builder_test.go

@@ -11,6 +11,7 @@ import (
 	qbtypes "github.com/SigNoz/signoz/pkg/types/querybuildertypes/querybuildertypesv5"
 	"github.com/SigNoz/signoz/pkg/types/telemetrytypes"
 	"github.com/SigNoz/signoz/pkg/types/telemetrytypes/telemetrytypestest"
+	"github.com/SigNoz/signoz/pkg/valuer"
 	"github.com/stretchr/testify/require"
 )
 
@@ -213,7 +214,7 @@ func TestStatementBuilder(t *testing.T) {
 
 	for _, testCase := range testCases {
 		t.Run(testCase.name, func(t *testing.T) {
-			q, err := statementBuilder.Build(ctx, 1747947419000, 1747983448000, testCase.requestType, testCase.query, nil)
+			q, err := statementBuilder.Build(ctx, valuer.UUID{}, 1747947419000, 1747983448000, testCase.requestType, testCase.query, nil)
 			if testCase.expectedErr != nil {
 				require.Error(t, err)
 				require.Contains(t, err.Error(), testCase.expectedErr.Error())

pkg/telemetrylogs/condition_builder.go

@@ -372,6 +372,7 @@ func (c *conditionBuilder) conditionFor(
 
 func (c *conditionBuilder) ConditionFor(
 	ctx context.Context,
+	orgID valuer.UUID,
 	startNs uint64,
 	endNs uint64,
 	key *telemetrytypes.TelemetryFieldKey,

pkg/telemetrylogs/condition_builder_test.go

@@ -8,6 +8,7 @@ import (
 	"github.com/SigNoz/signoz/pkg/flagger/flaggertest"
 	qbtypes "github.com/SigNoz/signoz/pkg/types/querybuildertypes/querybuildertypesv5"
 	"github.com/SigNoz/signoz/pkg/types/telemetrytypes"
+	"github.com/SigNoz/signoz/pkg/valuer"
 	"github.com/huandu/go-sqlbuilder"
 	"github.com/stretchr/testify/assert"
 	"github.com/stretchr/testify/require"
@@ -131,7 +132,7 @@ func TestExistsConditionForWithEvolutions(t *testing.T) {
 	for _, tc := range testCases {
 		sb := sqlbuilder.NewSelectBuilder()
 		t.Run(tc.name, func(t *testing.T) {
-			cond, _, err := conditionBuilder.ConditionFor(ctx, tc.startTs, tc.endTs, &tc.key, []*telemetrytypes.TelemetryFieldKey{&tc.key}, tc.operator, tc.value, sb)
+			cond, _, err := conditionBuilder.ConditionFor(ctx, valuer.UUID{}, tc.startTs, tc.endTs, &tc.key, []*telemetrytypes.TelemetryFieldKey{&tc.key}, tc.operator, tc.value, sb)
 			sb.Where(cond...)
 
 			if tc.expectedError != nil {
@@ -522,7 +523,7 @@ func TestConditionFor(t *testing.T) {
 		sb := sqlbuilder.NewSelectBuilder()
 		t.Run(tc.name, func(t *testing.T) {
 			tc.key.Evolutions = tc.evolutions
-			cond, _, err := conditionBuilder.ConditionFor(ctx, 0, 0, &tc.key, []*telemetrytypes.TelemetryFieldKey{&tc.key}, tc.operator, tc.value, sb)
+			cond, _, err := conditionBuilder.ConditionFor(ctx, valuer.UUID{}, 0, 0, &tc.key, []*telemetrytypes.TelemetryFieldKey{&tc.key}, tc.operator, tc.value, sb)
 			sb.Where(cond...)
 
 			if tc.expectedError != nil {

pkg/telemetrylogs/field_mapper.go

@@ -242,6 +242,7 @@ func (m *fieldMapper) ColumnFor(ctx context.Context, _, _ uint64, key *telemetry
 
 func (m *fieldMapper) ColumnExpressionFor(
 	ctx context.Context,
+	_ valuer.UUID,
 	tsStart, tsEnd uint64,
 	field *telemetrytypes.TelemetryFieldKey,
 	keys map[string][]*telemetrytypes.TelemetryFieldKey,
@@ -285,6 +286,12 @@ func (m *fieldMapper) ColumnExpressionFor(
 	return fmt.Sprintf("%s AS `%s`", sqlbuilder.Escape(fieldExpression), field.Name), nil
 }
 
+// CandidateKeys returns nil: logs has no synthesize-on-unknown-key fallback, so an
+// unknown key stays unresolved and the caller errors.
+func (m *fieldMapper) CandidateKeys(_ context.Context, _ valuer.UUID, _ *telemetrytypes.TelemetryFieldKey, _ any, _ map[string][]*telemetrytypes.TelemetryFieldKey) []*telemetrytypes.TelemetryFieldKey {
+	return nil
+}
+
 // buildFieldForJSON builds the field expression for body JSON fields using arrayConcat pattern.
 func (m *fieldMapper) buildFieldForJSON(key *telemetrytypes.TelemetryFieldKey) (string, error) {
 	plan := key.JSONPlan

pkg/telemetrylogs/json_stmt_builder_test.go

@@ -14,6 +14,7 @@ import (
 	qbtypes "github.com/SigNoz/signoz/pkg/types/querybuildertypes/querybuildertypesv5"
 	"github.com/SigNoz/signoz/pkg/types/telemetrytypes"
 	"github.com/SigNoz/signoz/pkg/types/telemetrytypes/telemetrytypestest"
+	"github.com/SigNoz/signoz/pkg/valuer"
 	"github.com/stretchr/testify/require"
 )
 
@@ -94,7 +95,7 @@ func TestJSONStmtBuilder_TimeSeries(t *testing.T) {
 	for _, c := range cases {
 		t.Run(c.name, func(t *testing.T) {
 
-			q, err := statementBuilder.Build(context.Background(), 1747947419000, 1747983448000, c.requestType, c.query, nil)
+			q, err := statementBuilder.Build(context.Background(), valuer.UUID{}, 1747947419000, 1747983448000, c.requestType, c.query, nil)
 
 			if c.expectedErrContains != "" {
 				require.Error(t, err)
@@ -155,7 +156,7 @@ func TestStmtBuilderTimeSeriesBodyGroupByPromoted(t *testing.T) {
 
 	for _, c := range cases {
 		t.Run(c.name, func(t *testing.T) {
-			q, err := statementBuilder.Build(context.Background(), 1747947419000, 1747983448000, c.requestType, c.query, nil)
+			q, err := statementBuilder.Build(context.Background(), valuer.UUID{}, 1747947419000, 1747983448000, c.requestType, c.query, nil)
 			if c.expectedErrContains != "" {
 				require.Error(t, err)
 				require.Contains(t, err.Error(), c.expectedErrContains)
@@ -308,7 +309,7 @@ func TestJSONStmtBuilder_PrimitivePaths(t *testing.T) {
 
 	for _, c := range cases {
 		t.Run(c.name, func(t *testing.T) {
-			q, err := statementBuilder.Build(context.Background(), 1747947419000, 1747983448000, qbtypes.RequestTypeRaw,
+			q, err := statementBuilder.Build(context.Background(), valuer.UUID{}, 1747947419000, 1747983448000, qbtypes.RequestTypeRaw,
 				qbtypes.QueryBuilderQuery[qbtypes.LogAggregation]{
 					Signal: telemetrytypes.SignalLogs,
 					Filter: &qbtypes.Filter{Expression: c.filter},
@@ -477,7 +478,7 @@ func TestStatementBuilderListQueryBodyPromoted(t *testing.T) {
 	for _, c := range cases {
 		t.Run(c.name, func(t *testing.T) {
 
-			q, err := statementBuilder.Build(context.Background(), 1747947419000, 1747983448000, c.requestType, c.query, nil)
+			q, err := statementBuilder.Build(context.Background(), valuer.UUID{}, 1747947419000, 1747983448000, c.requestType, c.query, nil)
 
 			if c.expectedErr != nil {
 				require.Error(t, err)
@@ -779,7 +780,7 @@ func TestJSONStmtBuilder_ArrayPaths(t *testing.T) {
 
 	for _, c := range cases {
 		t.Run(c.name, func(t *testing.T) {
-			q, err := statementBuilder.Build(context.Background(), 1747947419000, 1747983448000, qbtypes.RequestTypeRaw,
+			q, err := statementBuilder.Build(context.Background(), valuer.UUID{}, 1747947419000, 1747983448000, qbtypes.RequestTypeRaw,
 				qbtypes.QueryBuilderQuery[qbtypes.LogAggregation]{
 					Signal: telemetrytypes.SignalLogs,
 					Filter: &qbtypes.Filter{Expression: c.filter},
@@ -904,7 +905,7 @@ func TestJSONStmtBuilder_IndexedPaths(t *testing.T) {
 
 	for _, c := range cases {
 		t.Run(c.name, func(t *testing.T) {
-			q, err := statementBuilder.Build(context.Background(), 1747947419000, 1747983448000, qbtypes.RequestTypeRaw, c.query, nil)
+			q, err := statementBuilder.Build(context.Background(), valuer.UUID{}, 1747947419000, 1747983448000, qbtypes.RequestTypeRaw, c.query, nil)
 			if c.expectedErr != nil {
 				require.Error(t, err)
 				require.Contains(t, err.Error(), c.expectedErr.Error())
@@ -991,7 +992,7 @@ func TestJSONStmtBuilder_SelectField(t *testing.T) {
 
 	for _, c := range cases {
 		t.Run(c.name, func(t *testing.T) {
-			q, err := statementBuilder.Build(context.Background(), 1747947419000, 1747983448000, c.requestType, c.query, nil)
+			q, err := statementBuilder.Build(context.Background(), valuer.UUID{}, 1747947419000, 1747983448000, c.requestType, c.query, nil)
 			if c.expectedErrContains != "" {
 				require.Error(t, err)
 				require.Contains(t, err.Error(), c.expectedErrContains)
@@ -1068,7 +1069,7 @@ func TestJSONStmtBuilder_OrderBy(t *testing.T) {
 
 	for _, c := range cases {
 		t.Run(c.name, func(t *testing.T) {
-			q, err := statementBuilder.Build(context.Background(), 1747947419000, 1747983448000, c.requestType, c.query, nil)
+			q, err := statementBuilder.Build(context.Background(), valuer.UUID{}, 1747947419000, 1747983448000, c.requestType, c.query, nil)
 			if c.expectedErrContains != "" {
 				require.Error(t, err)
 				require.Contains(t, err.Error(), c.expectedErrContains)
@@ -1131,7 +1132,7 @@ func TestResourceAggrAndGroupBy_WithJSONEnabled(t *testing.T) {
 
 	for _, c := range cases {
 		t.Run(c.name, func(t *testing.T) {
-			q, err := statementBuilder.Build(context.Background(), 1747947419000, 1747983448000, c.requestType, c.query, nil)
+			q, err := statementBuilder.Build(context.Background(), valuer.UUID{}, 1747947419000, 1747983448000, c.requestType, c.query, nil)
 			if c.expectedErrContains != "" {
 				require.Error(t, err)
 				require.Contains(t, err.Error(), c.expectedErrContains)

pkg/telemetrylogs/statement_builder.go

@@ -80,6 +80,7 @@ func NewLogQueryStatementBuilder(
 // Build builds a SQL query for logs based on the given parameters.
 func (b *logQueryStatementBuilder) Build(
 	ctx context.Context,
+	orgID valuer.UUID,
 	start uint64,
 	end uint64,
 	requestType qbtypes.RequestType,
@@ -106,11 +107,11 @@ func (b *logQueryStatementBuilder) Build(
 	var stmt *qbtypes.Statement
 	switch requestType {
 	case qbtypes.RequestTypeRaw, qbtypes.RequestTypeRawStream:
-		stmt, err = b.buildListQuery(ctx, q, query, start, end, keys, variables)
+		stmt, err = b.buildListQuery(ctx, orgID, q, query, start, end, keys, variables)
 	case qbtypes.RequestTypeTimeSeries:
-		stmt, err = b.buildTimeSeriesQuery(ctx, q, query, start, end, keys, variables)
+		stmt, err = b.buildTimeSeriesQuery(ctx, orgID, q, query, start, end, keys, variables)
 	case qbtypes.RequestTypeScalar:
-		stmt, err = b.buildScalarQuery(ctx, q, query, start, end, keys, false, variables)
+		stmt, err = b.buildScalarQuery(ctx, orgID, q, query, start, end, keys, false, variables)
 	default:
 		return nil, errors.NewInvalidInputf(errors.CodeInvalidInput, "unsupported request type: %s", requestType)
 	}
@@ -264,6 +265,7 @@ func (b *logQueryStatementBuilder) adjustKey(key *telemetrytypes.TelemetryFieldK
 // buildListQuery builds a query for list panel type.
 func (b *logQueryStatementBuilder) buildListQuery(
 	ctx context.Context,
+	orgID valuer.UUID,
 	sb *sqlbuilder.SelectBuilder,
 	query qbtypes.QueryBuilderQuery[qbtypes.LogAggregation],
 	start, end uint64,
@@ -278,7 +280,7 @@ func (b *logQueryStatementBuilder) buildListQuery(
 		bodyJSONEnabled = b.fl.BooleanOrEmpty(ctx, flagger.FeatureUseJSONBody, featuretypes.NewFlaggerEvaluationContext(valuer.UUID{}))
 	)
 
-	frag, args, skipResourceFilter, err := b.maybeAttachResourceFilter(ctx, sb, query, start, end, variables)
+	frag, args, skipResourceFilter, err := b.maybeAttachResourceFilter(ctx, orgID, sb, query, start, end, variables)
 	if err != nil {
 		return nil, err
 	}
@@ -314,7 +316,7 @@ func (b *logQueryStatementBuilder) buildListQuery(
 			}
 
 			// get column expression for the field - use array index directly to avoid pointer to loop variable
-			colExpr, err := b.fm.ColumnExpressionFor(ctx, start, end, &query.SelectFields[index], keys)
+			colExpr, err := b.fm.ColumnExpressionFor(ctx, orgID, start, end, &query.SelectFields[index], keys)
 			if err != nil {
 				return nil, err
 			}
@@ -324,7 +326,7 @@ func (b *logQueryStatementBuilder) buildListQuery(
 
 	sb.From(fmt.Sprintf("%s.%s", DBName, LogsV2TableName))
 	// Add filter conditions
-	preparedWhereClause, err := b.addFilterCondition(ctx, sb, start, end, query, keys, variables, skipResourceFilter)
+	preparedWhereClause, err := b.addFilterCondition(ctx, orgID, sb, start, end, query, keys, variables, skipResourceFilter)
 
 	if err != nil {
 		return nil, err
@@ -333,7 +335,7 @@ func (b *logQueryStatementBuilder) buildListQuery(
 	// Add order by
 	for _, orderBy := range query.Order {
 
-		colExpr, err := b.fm.ColumnExpressionFor(ctx, start, end, &orderBy.Key.TelemetryFieldKey, keys)
+		colExpr, err := b.fm.ColumnExpressionFor(ctx, orgID, start, end, &orderBy.Key.TelemetryFieldKey, keys)
 		if err != nil {
 			return nil, err
 		}
@@ -368,6 +370,7 @@ func (b *logQueryStatementBuilder) buildListQuery(
 
 func (b *logQueryStatementBuilder) buildTimeSeriesQuery(
 	ctx context.Context,
+	orgID valuer.UUID,
 	sb *sqlbuilder.SelectBuilder,
 	query qbtypes.QueryBuilderQuery[qbtypes.LogAggregation],
 	start, end uint64,
@@ -382,7 +385,7 @@ func (b *logQueryStatementBuilder) buildTimeSeriesQuery(
 		bodyJSONEnabled = b.fl.BooleanOrEmpty(ctx, flagger.FeatureUseJSONBody, featuretypes.NewFlaggerEvaluationContext(valuer.UUID{}))
 	)
 
-	frag, args, skipResourceFilter, err := b.maybeAttachResourceFilter(ctx, sb, query, start, end, variables)
+	frag, args, skipResourceFilter, err := b.maybeAttachResourceFilter(ctx, orgID, sb, query, start, end, variables)
 	if err != nil {
 		return nil, err
 	}
@@ -401,7 +404,7 @@ func (b *logQueryStatementBuilder) buildTimeSeriesQuery(
 	// Keep original column expressions so we can build the tuple
 	fieldNames := make([]string, 0, len(query.GroupBy))
 	for _, gb := range query.GroupBy {
-		expr, args, err := querybuilder.CollisionHandledFinalExpr(ctx, start, end, &gb.TelemetryFieldKey, b.fm, b.cb, keys, telemetrytypes.FieldDataTypeString, b.jsonKeyToKey, bodyJSONEnabled)
+		expr, args, err := querybuilder.CollisionHandledFinalExpr(ctx, orgID, start, end, &gb.TelemetryFieldKey, b.fm, b.cb, keys, telemetrytypes.FieldDataTypeString, b.jsonKeyToKey, bodyJSONEnabled)
 		if err != nil {
 			return nil, err
 		}
@@ -416,7 +419,7 @@ func (b *logQueryStatementBuilder) buildTimeSeriesQuery(
 	allAggChArgs := make([]any, 0)
 	for i, agg := range query.Aggregations {
 		rewritten, chArgs, err := b.aggExprRewriter.Rewrite(
-			ctx, start, end, agg.Expression,
+			ctx, orgID, start, end, agg.Expression,
 			uint64(query.StepInterval.Seconds()),
 			keys,
 		)
@@ -430,7 +433,7 @@ func (b *logQueryStatementBuilder) buildTimeSeriesQuery(
 	// Add FROM clause
 	sb.From(fmt.Sprintf("%s.%s", DBName, LogsV2TableName))
 
-	preparedWhereClause, err := b.addFilterCondition(ctx, sb, start, end, query, keys, variables, skipResourceFilter)
+	preparedWhereClause, err := b.addFilterCondition(ctx, orgID, sb, start, end, query, keys, variables, skipResourceFilter)
 
 	if err != nil {
 		return nil, err
@@ -442,7 +445,7 @@ func (b *logQueryStatementBuilder) buildTimeSeriesQuery(
 	if query.Limit > 0 && len(query.GroupBy) > 0 {
 		// build the scalar “top/bottom-N” query in its own builder.
 		cteSB := sqlbuilder.NewSelectBuilder()
-		cteStmt, err := b.buildScalarQuery(ctx, cteSB, query, start, end, keys, true, variables)
+		cteStmt, err := b.buildScalarQuery(ctx, orgID, cteSB, query, start, end, keys, true, variables)
 		if err != nil {
 			return nil, err
 		}
@@ -527,6 +530,7 @@ func (b *logQueryStatementBuilder) buildTimeSeriesQuery(
 // buildScalarQuery builds a query for scalar panel type.
 func (b *logQueryStatementBuilder) buildScalarQuery(
 	ctx context.Context,
+	orgID valuer.UUID,
 	sb *sqlbuilder.SelectBuilder,
 	query qbtypes.QueryBuilderQuery[qbtypes.LogAggregation],
 	start, end uint64,
@@ -542,7 +546,7 @@ func (b *logQueryStatementBuilder) buildScalarQuery(
 		bodyJSONEnabled = b.fl.BooleanOrEmpty(ctx, flagger.FeatureUseJSONBody, featuretypes.NewFlaggerEvaluationContext(valuer.UUID{}))
 	)
 
-	frag, args, skipResourceFilter, err := b.maybeAttachResourceFilter(ctx, sb, query, start, end, variables)
+	frag, args, skipResourceFilter, err := b.maybeAttachResourceFilter(ctx, orgID, sb, query, start, end, variables)
 	if err != nil {
 		return nil, err
 	}
@@ -556,7 +560,7 @@ func (b *logQueryStatementBuilder) buildScalarQuery(
 	var allGroupByArgs []any
 
 	for _, gb := range query.GroupBy {
-		expr, args, err := querybuilder.CollisionHandledFinalExpr(ctx, start, end, &gb.TelemetryFieldKey, b.fm, b.cb, keys, telemetrytypes.FieldDataTypeString, b.jsonKeyToKey, bodyJSONEnabled)
+		expr, args, err := querybuilder.CollisionHandledFinalExpr(ctx, orgID, start, end, &gb.TelemetryFieldKey, b.fm, b.cb, keys, telemetrytypes.FieldDataTypeString, b.jsonKeyToKey, bodyJSONEnabled)
 		if err != nil {
 			return nil, err
 		}
@@ -574,7 +578,7 @@ func (b *logQueryStatementBuilder) buildScalarQuery(
 		for idx := range query.Aggregations {
 			aggExpr := query.Aggregations[idx]
 			rewritten, chArgs, err := b.aggExprRewriter.Rewrite(
-				ctx, start, end, aggExpr.Expression,
+				ctx, orgID, start, end, aggExpr.Expression,
 				rateInterval,
 				keys,
 			)
@@ -589,7 +593,7 @@ func (b *logQueryStatementBuilder) buildScalarQuery(
 	sb.From(fmt.Sprintf("%s.%s", DBName, LogsV2TableName))
 
 	// Add filter conditions
-	preparedWhereClause, err := b.addFilterCondition(ctx, sb, start, end, query, keys, variables, skipResourceFilter)
+	preparedWhereClause, err := b.addFilterCondition(ctx, orgID, sb, start, end, query, keys, variables, skipResourceFilter)
 
 	if err != nil {
 		return nil, err
@@ -648,6 +652,7 @@ func (b *logQueryStatementBuilder) buildScalarQuery(
 // buildFilterCondition builds SQL condition from filter expression.
 func (b *logQueryStatementBuilder) addFilterCondition(
 	ctx context.Context,
+	orgID valuer.UUID,
 	sb *sqlbuilder.SelectBuilder,
 	start, end uint64,
 	query qbtypes.QueryBuilderQuery[qbtypes.LogAggregation],
@@ -663,6 +668,7 @@ func (b *logQueryStatementBuilder) addFilterCondition(
 		// add filter expression
 		preparedWhereClause, err = querybuilder.PrepareWhereClause(query.Filter.Expression, querybuilder.FilterExprVisitorOpts{
 			Context:            ctx,
+			OrgID:              orgID,
 			Logger:             b.logger,
 			FieldMapper:        b.fm,
 			ConditionBuilder:   b.cb,
@@ -713,6 +719,7 @@ func aggOrderBy(k qbtypes.OrderBy, q qbtypes.QueryBuilderQuery[qbtypes.LogAggreg
 
 func (b *logQueryStatementBuilder) maybeAttachResourceFilter(
 	ctx context.Context,
+	orgID valuer.UUID,
 	sb *sqlbuilder.SelectBuilder,
 	query qbtypes.QueryBuilderQuery[qbtypes.LogAggregation],
 	start, end uint64,
@@ -720,7 +727,7 @@ func (b *logQueryStatementBuilder) maybeAttachResourceFilter(
 ) (cteSQL string, cteArgs []any, skipResourceFilter bool, err error) {
 
 	if b.skipResourceFingerprintEnabled {
-		decision, err := b.resourceFilterResolver.Resolve(ctx, query, start, end, variables)
+		decision, err := b.resourceFilterResolver.Resolve(ctx, orgID, query, start, end, variables)
 		if err != nil {
 			return "", nil, true, err
 		}
@@ -733,7 +740,7 @@ func (b *logQueryStatementBuilder) maybeAttachResourceFilter(
 	}
 
 	stmt, err := b.resourceFilterResolver.StatementBuilder().Build(
-		ctx, start, end, qbtypes.RequestTypeRaw, query, variables,
+		ctx, orgID, start, end, qbtypes.RequestTypeRaw, query, variables,
 	)
 	if err != nil {
 		return "", nil, true, err

pkg/telemetrylogs/stmt_builder_test.go

@@ -16,6 +16,7 @@ import (
 	qbtypes "github.com/SigNoz/signoz/pkg/types/querybuildertypes/querybuildertypesv5"
 	"github.com/SigNoz/signoz/pkg/types/telemetrytypes"
 	"github.com/SigNoz/signoz/pkg/types/telemetrytypes/telemetrytypestest"
+	"github.com/SigNoz/signoz/pkg/valuer"
 	"github.com/stretchr/testify/require"
 )
 
@@ -237,7 +238,7 @@ func TestStatementBuilderTimeSeries(t *testing.T) {
 	for _, c := range cases {
 		t.Run(c.name, func(t *testing.T) {
 
-			q, err := statementBuilder.Build(ctx, c.startTs, c.endTs, c.requestType, c.query, nil)
+			q, err := statementBuilder.Build(ctx, valuer.UUID{}, c.startTs, c.endTs, c.requestType, c.query, nil)
 
 			if c.expectedErr != nil {
 				require.Error(t, err)
@@ -381,7 +382,7 @@ func TestStatementBuilderListQuery(t *testing.T) {
 	for _, c := range cases {
 		t.Run(c.name, func(t *testing.T) {
 
-			q, err := statementBuilder.Build(ctx, 1747947419000, 1747983448000, c.requestType, c.query, c.variables)
+			q, err := statementBuilder.Build(ctx, valuer.UUID{}, 1747947419000, 1747983448000, c.requestType, c.query, c.variables)
 
 			if c.expectedErr != nil {
 				require.Error(t, err)
@@ -530,7 +531,7 @@ func TestStatementBuilderListQueryResourceTests(t *testing.T) {
 	for _, c := range cases {
 		t.Run(c.name, func(t *testing.T) {
 
-			q, err := statementBuilder.Build(ctx, 1747947419000, 1747983448000, c.requestType, c.query, nil)
+			q, err := statementBuilder.Build(ctx, valuer.UUID{}, 1747947419000, 1747983448000, c.requestType, c.query, nil)
 
 			if c.expectedErr != nil {
 				require.Error(t, err)
@@ -609,7 +610,7 @@ func TestStatementBuilderTimeSeriesBodyGroupBy(t *testing.T) {
 	for _, c := range cases {
 		t.Run(c.name, func(t *testing.T) {
 
-			q, err := statementBuilder.Build(ctx, 1747947419000, 1747983448000, c.requestType, c.query, nil)
+			q, err := statementBuilder.Build(ctx, valuer.UUID{}, 1747947419000, 1747983448000, c.requestType, c.query, nil)
 
 			if c.expectedErrContains != "" {
 				require.Error(t, err)
@@ -707,7 +708,7 @@ func TestStatementBuilderListQueryServiceCollision(t *testing.T) {
 	for _, c := range cases {
 		t.Run(c.name, func(t *testing.T) {
 
-			q, err := statementBuilder.Build(ctx, 1747947419000, 1747983448000, c.requestType, c.query, nil)
+			q, err := statementBuilder.Build(ctx, valuer.UUID{}, 1747947419000, 1747983448000, c.requestType, c.query, nil)
 
 			if c.expectedErr != nil {
 				require.Error(t, err)
@@ -1079,7 +1080,7 @@ func TestStmtBuilderBodyField(t *testing.T) {
 				100000,
 			)
 
-			q, err := statementBuilder.Build(context.Background(), 1747947419000, 1747983448000, c.requestType, c.query, nil)
+			q, err := statementBuilder.Build(context.Background(), valuer.UUID{}, 1747947419000, 1747983448000, c.requestType, c.query, nil)
 			if c.expectedErr != nil {
 				require.Error(t, err)
 				require.Contains(t, err.Error(), c.expectedErr.Error())
@@ -1181,7 +1182,7 @@ func TestStmtBuilderBodyFullTextSearch(t *testing.T) {
 				100000,
 			)
 
-			q, err := statementBuilder.Build(context.Background(), 1747947419000, 1747983448000, c.requestType, c.query, nil)
+			q, err := statementBuilder.Build(context.Background(), valuer.UUID{}, 1747947419000, 1747983448000, c.requestType, c.query, nil)
 			if c.expectedErr != nil {
 				require.Error(t, err)
 				require.Contains(t, err.Error(), c.expectedErr.Error())
@@ -1220,7 +1221,7 @@ func TestSkipResourceFingerprintLogs(t *testing.T) {
 	t.Run("disabled uses the legacy CTE", func(t *testing.T) {
 		sb := newSkipResourceFingerprintLogsBuilder(t, nil, false, threshold)
 
-		stmt, err := sb.Build(context.Background(), startMs, endMs, qbtypes.RequestTypeRaw, query, nil)
+		stmt, err := sb.Build(context.Background(), valuer.UUID{}, startMs, endMs, qbtypes.RequestTypeRaw, query, nil)
 		require.NoError(t, err)
 		require.Contains(t, stmt.Query, "__resource_filter AS (SELECT fingerprint")
 		require.Contains(t, stmt.Query, "resource_fingerprint GLOBAL IN (SELECT fingerprint FROM __resource_filter)")
@@ -1237,7 +1238,7 @@ func TestSkipResourceFingerprintLogs(t *testing.T) {
 
 		sb := newSkipResourceFingerprintLogsBuilder(t, mockStore, true, threshold)
 
-		stmt, err := sb.Build(context.Background(), startMs, endMs, qbtypes.RequestTypeRaw, query, nil)
+		stmt, err := sb.Build(context.Background(), valuer.UUID{}, startMs, endMs, qbtypes.RequestTypeRaw, query, nil)
 		require.NoError(t, err)
 
 		require.Contains(t, stmt.Query, "__resource_filter AS (SELECT fingerprint")
@@ -1257,7 +1258,7 @@ func TestSkipResourceFingerprintLogs(t *testing.T) {
 
 		sb := newSkipResourceFingerprintLogsBuilder(t, mockStore, true, threshold)
 
-		stmt, err := sb.Build(context.Background(), startMs, endMs, qbtypes.RequestTypeRaw, query, nil)
+		stmt, err := sb.Build(context.Background(), valuer.UUID{}, startMs, endMs, qbtypes.RequestTypeRaw, query, nil)
 		require.NoError(t, err)
 
 		require.NotContains(t, stmt.Query, "__resource_filter AS")

pkg/telemetrymetadata/condition_builder.go

@@ -8,6 +8,7 @@ import (
 	"github.com/SigNoz/signoz/pkg/querybuilder"
 	qbtypes "github.com/SigNoz/signoz/pkg/types/querybuildertypes/querybuildertypesv5"
 	"github.com/SigNoz/signoz/pkg/types/telemetrytypes"
+	"github.com/SigNoz/signoz/pkg/valuer"
 	"github.com/huandu/go-sqlbuilder"
 )
 
@@ -21,6 +22,7 @@ func NewConditionBuilder(fm qbtypes.FieldMapper) *conditionBuilder {
 
 func (c *conditionBuilder) ConditionFor(
 	ctx context.Context,
+	orgID valuer.UUID,
 	tsStart, tsEnd uint64,
 	key *telemetrytypes.TelemetryFieldKey,
 	fieldKeysForName []*telemetrytypes.TelemetryFieldKey,

pkg/telemetrymetadata/condition_builder_test.go

@@ -6,6 +6,7 @@ import (
 
 	qbtypes "github.com/SigNoz/signoz/pkg/types/querybuildertypes/querybuildertypesv5"
 	"github.com/SigNoz/signoz/pkg/types/telemetrytypes"
+	"github.com/SigNoz/signoz/pkg/valuer"
 	"github.com/huandu/go-sqlbuilder"
 	"github.com/stretchr/testify/assert"
 	"github.com/stretchr/testify/require"
@@ -53,7 +54,7 @@ func TestConditionFor(t *testing.T) {
 	for _, tc := range testCases {
 		sb := sqlbuilder.NewSelectBuilder()
 		t.Run(tc.name, func(t *testing.T) {
-			cond, _, err := conditionBuilder.ConditionFor(ctx, 0, 0, &tc.key, []*telemetrytypes.TelemetryFieldKey{&tc.key}, tc.operator, tc.value, sb)
+			cond, _, err := conditionBuilder.ConditionFor(ctx, valuer.UUID{}, 0, 0, &tc.key, []*telemetrytypes.TelemetryFieldKey{&tc.key}, tc.operator, tc.value, sb)
 			sb.Where(cond...)
 
 			if tc.expectedError != nil {

pkg/telemetrymetadata/field_mapper.go

@@ -9,6 +9,7 @@ import (
 	"github.com/SigNoz/signoz/pkg/errors"
 	qbtypes "github.com/SigNoz/signoz/pkg/types/querybuildertypes/querybuildertypesv5"
 	"github.com/SigNoz/signoz/pkg/types/telemetrytypes"
+	"github.com/SigNoz/signoz/pkg/valuer"
 	"github.com/huandu/go-sqlbuilder"
 	"golang.org/x/exp/maps"
 )
@@ -29,6 +30,12 @@ var (
 type fieldMapper struct {
 }
 
+// CandidateKeys returns nil: this mapper has no attribute-map fallback, so a context-missing
+// key stays unresolved and the caller errors.
+func (m *fieldMapper) CandidateKeys(_ context.Context, _ valuer.UUID, _ *telemetrytypes.TelemetryFieldKey, _ any, _ map[string][]*telemetrytypes.TelemetryFieldKey) []*telemetrytypes.TelemetryFieldKey {
+	return nil
+}
+
 func NewFieldMapper() qbtypes.FieldMapper {
 	return &fieldMapper{}
 }
@@ -69,6 +76,7 @@ func (m *fieldMapper) FieldFor(ctx context.Context, startNs, endNs uint64, key *
 
 func (m *fieldMapper) ColumnExpressionFor(
 	ctx context.Context,
+	_ valuer.UUID,
 	startNs, endNs uint64,
 	field *telemetrytypes.TelemetryFieldKey,
 	keys map[string][]*telemetrytypes.TelemetryFieldKey,

pkg/telemetrymetadata/metadata.go

@@ -1190,6 +1190,27 @@ func enrichWithIntrinsicMetricKeys(keys map[string][]*telemetrytypes.TelemetryFi
 	return keys
 }
 
+// enrichWithGenAIKeys adds keys that can be queried for GenAI signals, even though they have not been ingested yet.
+func enrichWithGenAIKeys(keys map[string][]*telemetrytypes.TelemetryFieldKey, selectors []*telemetrytypes.FieldKeySelector) map[string][]*telemetrytypes.TelemetryFieldKey {
+	for _, selector := range selectors {
+		if selector.Signal != telemetrytypes.SignalTraces && selector.Signal != telemetrytypes.SignalUnspecified {
+			continue
+		}
+		for name, def := range telemetrytypes.GenAIFieldDefinitions {
+			if len(keys[name]) > 0 {
+				continue // already resolved from ingested data
+			}
+			if !selectorMatchesIntrinsicField(selector, def) {
+				continue
+			}
+			keyCopy := def
+			keys[name] = []*telemetrytypes.TelemetryFieldKey{&keyCopy}
+		}
+	}
+
+	return keys
+}
+
 func selectorMatchesIntrinsicField(selector *telemetrytypes.FieldKeySelector, definition telemetrytypes.TelemetryFieldKey) bool {
 	if selector.FieldContext != telemetrytypes.FieldContextUnspecified && selector.FieldContext != definition.FieldContext {
 		return false
@@ -1275,6 +1296,7 @@ func (t *telemetryMetaStore) GetKeys(ctx context.Context, fieldKeySelector *tele
 
 	applyBackwardCompatibleKeys(mapOfKeys)
 	mapOfKeys = enrichWithIntrinsicMetricKeys(mapOfKeys, selectors)
+	mapOfKeys = enrichWithGenAIKeys(mapOfKeys, selectors)
 
 	return mapOfKeys, complete, nil
 }
@@ -1353,6 +1375,7 @@ func (t *telemetryMetaStore) GetKeysMulti(ctx context.Context, fieldKeySelectors
 
 	applyBackwardCompatibleKeys(mapOfKeys)
 	mapOfKeys = enrichWithIntrinsicMetricKeys(mapOfKeys, fieldKeySelectors)
+	mapOfKeys = enrichWithGenAIKeys(mapOfKeys, fieldKeySelectors)
 
 	return mapOfKeys, complete, nil
 }
@@ -1446,20 +1469,20 @@ func (t *telemetryMetaStore) getRelatedValues(ctx context.Context, fieldValueSel
 
 			// search on attributes
 			key.FieldContext = telemetrytypes.FieldContextAttribute
-			attrConds, _, err := t.conditionBuilder.ConditionFor(ctx, 0, 0, key, []*telemetrytypes.TelemetryFieldKey{key}, qbtypes.FilterOperatorContains, fieldValueSelector.Value, sb)
+			attrConds, _, err := t.conditionBuilder.ConditionFor(ctx, valuer.UUID{}, 0, 0, key, []*telemetrytypes.TelemetryFieldKey{key}, qbtypes.FilterOperatorContains, fieldValueSelector.Value, sb)
 			if err == nil {
 				conds = append(conds, attrConds...)
 			}
 
 			// search on resource
 			key.FieldContext = telemetrytypes.FieldContextResource
-			resourceConds, _, err := t.conditionBuilder.ConditionFor(ctx, 0, 0, key, []*telemetrytypes.TelemetryFieldKey{key}, qbtypes.FilterOperatorContains, fieldValueSelector.Value, sb)
+			resourceConds, _, err := t.conditionBuilder.ConditionFor(ctx, valuer.UUID{}, 0, 0, key, []*telemetrytypes.TelemetryFieldKey{key}, qbtypes.FilterOperatorContains, fieldValueSelector.Value, sb)
 			if err == nil {
 				conds = append(conds, resourceConds...)
 			}
 			key.FieldContext = origContext
 		} else {
-			keyConds, _, err := t.conditionBuilder.ConditionFor(ctx, 0, 0, key, []*telemetrytypes.TelemetryFieldKey{key}, qbtypes.FilterOperatorContains, fieldValueSelector.Value, sb)
+			keyConds, _, err := t.conditionBuilder.ConditionFor(ctx, valuer.UUID{}, 0, 0, key, []*telemetrytypes.TelemetryFieldKey{key}, qbtypes.FilterOperatorContains, fieldValueSelector.Value, sb)
 			if err == nil {
 				conds = append(conds, keyConds...)
 			}

pkg/telemetrymeter/statement_builder.go

@@ -12,6 +12,7 @@ import (
 	"github.com/SigNoz/signoz/pkg/types/metrictypes"
 	qbtypes "github.com/SigNoz/signoz/pkg/types/querybuildertypes/querybuildertypesv5"
 	"github.com/SigNoz/signoz/pkg/types/telemetrytypes"
+	"github.com/SigNoz/signoz/pkg/valuer"
 	"github.com/huandu/go-sqlbuilder"
 )
 
@@ -45,6 +46,7 @@ func NewMeterQueryStatementBuilder(
 
 func (b *meterQueryStatementBuilder) Build(
 	ctx context.Context,
+	orgID valuer.UUID,
 	start uint64,
 	end uint64,
 	_ qbtypes.RequestType,
@@ -59,11 +61,12 @@ func (b *meterQueryStatementBuilder) Build(
 
 	start, end = querybuilder.AdjustedMetricTimeRange(start, end, uint64(query.StepInterval.Seconds()), query)
 
-	return b.buildPipelineStatement(ctx, start, end, query, keys, variables)
+	return b.buildPipelineStatement(ctx, orgID, start, end, query, keys, variables)
 }
 
 func (b *meterQueryStatementBuilder) buildPipelineStatement(
 	ctx context.Context,
+	orgID valuer.UUID,
 	start, end uint64,
 	query qbtypes.QueryBuilderQuery[qbtypes.MetricAggregation],
 	keys map[string][]*telemetrytypes.TelemetryFieldKey,
@@ -76,7 +79,7 @@ func (b *meterQueryStatementBuilder) buildPipelineStatement(
 
 	if qbtypes.CanShortCircuitDelta(query.Aggregations[0]) {
 		// spatial_aggregation_cte directly for certain delta queries
-		if frag, args, err := b.buildTemporalAggDeltaFastPath(ctx, start, end, query, keys, variables); err != nil {
+		if frag, args, err := b.buildTemporalAggDeltaFastPath(ctx, orgID, start, end, query, keys, variables); err != nil {
 			return nil, err
 		} else if frag != "" {
 			cteFragments = append(cteFragments, frag)
@@ -84,7 +87,7 @@ func (b *meterQueryStatementBuilder) buildPipelineStatement(
 		}
 	} else {
 		// temporal_aggregation_cte
-		if frag, args, err := b.buildTemporalAggregationCTE(ctx, start, end, query, keys, variables); err != nil {
+		if frag, args, err := b.buildTemporalAggregationCTE(ctx, orgID, start, end, query, keys, variables); err != nil {
 			return nil, err
 		} else if frag != "" {
 			cteFragments = append(cteFragments, frag)
@@ -106,6 +109,7 @@ func (b *meterQueryStatementBuilder) buildPipelineStatement(
 
 func (b *meterQueryStatementBuilder) buildTemporalAggDeltaFastPath(
 	ctx context.Context,
+	orgID valuer.UUID,
 	start, end uint64,
 	query qbtypes.QueryBuilderQuery[qbtypes.MetricAggregation],
 	keys map[string][]*telemetrytypes.TelemetryFieldKey,
@@ -122,7 +126,7 @@ func (b *meterQueryStatementBuilder) buildTemporalAggDeltaFastPath(
 		stepSec,
 	))
 	for _, g := range query.GroupBy {
-		col, err := b.fm.ColumnExpressionFor(ctx, start, end, &g.TelemetryFieldKey, keys)
+		col, err := b.fm.ColumnExpressionFor(ctx, orgID, start, end, &g.TelemetryFieldKey, keys)
 		if err != nil {
 			return "", nil, err
 		}
@@ -148,6 +152,7 @@ func (b *meterQueryStatementBuilder) buildTemporalAggDeltaFastPath(
 	if query.Filter != nil && query.Filter.Expression != "" {
 		filterWhere, err = querybuilder.PrepareWhereClause(query.Filter.Expression, querybuilder.FilterExprVisitorOpts{
 			Context:          ctx,
+			OrgID:            orgID,
 			Logger:           b.logger,
 			FieldMapper:      b.fm,
 			ConditionBuilder: b.cb,
@@ -177,19 +182,21 @@ func (b *meterQueryStatementBuilder) buildTemporalAggDeltaFastPath(
 
 func (b *meterQueryStatementBuilder) buildTemporalAggregationCTE(
 	ctx context.Context,
+	orgID valuer.UUID,
 	start, end uint64,
 	query qbtypes.QueryBuilderQuery[qbtypes.MetricAggregation],
 	keys map[string][]*telemetrytypes.TelemetryFieldKey,
 	variables map[string]qbtypes.VariableItem,
 ) (string, []any, error) {
 	if query.Aggregations[0].Temporality == metrictypes.Delta {
-		return b.buildTemporalAggDelta(ctx, start, end, query, keys, variables)
+		return b.buildTemporalAggDelta(ctx, orgID, start, end, query, keys, variables)
 	}
-	return b.buildTemporalAggCumulativeOrUnspecified(ctx, start, end, query, keys, variables)
+	return b.buildTemporalAggCumulativeOrUnspecified(ctx, orgID, start, end, query, keys, variables)
 }
 
 func (b *meterQueryStatementBuilder) buildTemporalAggDelta(
 	ctx context.Context,
+	orgID valuer.UUID,
 	start, end uint64,
 	query qbtypes.QueryBuilderQuery[qbtypes.MetricAggregation],
 	keys map[string][]*telemetrytypes.TelemetryFieldKey,
@@ -208,7 +215,7 @@ func (b *meterQueryStatementBuilder) buildTemporalAggDelta(
 	))
 
 	for _, g := range query.GroupBy {
-		col, err := b.fm.ColumnExpressionFor(ctx, start, end, &g.TelemetryFieldKey, keys)
+		col, err := b.fm.ColumnExpressionFor(ctx, orgID, start, end, &g.TelemetryFieldKey, keys)
 		if err != nil {
 			return "", nil, err
 		}
@@ -237,6 +244,7 @@ func (b *meterQueryStatementBuilder) buildTemporalAggDelta(
 	if query.Filter != nil && query.Filter.Expression != "" {
 		filterWhere, err = querybuilder.PrepareWhereClause(query.Filter.Expression, querybuilder.FilterExprVisitorOpts{
 			Context:          ctx,
+			OrgID:            orgID,
 			Logger:           b.logger,
 			FieldMapper:      b.fm,
 			ConditionBuilder: b.cb,
@@ -268,6 +276,7 @@ func (b *meterQueryStatementBuilder) buildTemporalAggDelta(
 
 func (b *meterQueryStatementBuilder) buildTemporalAggCumulativeOrUnspecified(
 	ctx context.Context,
+	orgID valuer.UUID,
 	start, end uint64,
 	query qbtypes.QueryBuilderQuery[qbtypes.MetricAggregation],
 	keys map[string][]*telemetrytypes.TelemetryFieldKey,
@@ -284,7 +293,7 @@ func (b *meterQueryStatementBuilder) buildTemporalAggCumulativeOrUnspecified(
 		stepSec,
 	))
 	for _, g := range query.GroupBy {
-		col, err := b.fm.ColumnExpressionFor(ctx, start, end, &g.TelemetryFieldKey, keys)
+		col, err := b.fm.ColumnExpressionFor(ctx, orgID, start, end, &g.TelemetryFieldKey, keys)
 		if err != nil {
 			return "", nil, err
 		}
@@ -307,6 +316,7 @@ func (b *meterQueryStatementBuilder) buildTemporalAggCumulativeOrUnspecified(
 	if query.Filter != nil && query.Filter.Expression != "" {
 		filterWhere, err = querybuilder.PrepareWhereClause(query.Filter.Expression, querybuilder.FilterExprVisitorOpts{
 			Context:          ctx,
+			OrgID:            orgID,
 			Logger:           b.logger,
 			FieldMapper:      b.fm,
 			ConditionBuilder: b.cb,

pkg/telemetrymeter/stmt_builder_test.go

@@ -12,6 +12,7 @@ import (
 	qbtypes "github.com/SigNoz/signoz/pkg/types/querybuildertypes/querybuildertypesv5"
 	"github.com/SigNoz/signoz/pkg/types/telemetrytypes"
 	"github.com/SigNoz/signoz/pkg/types/telemetrytypes/telemetrytypestest"
+	"github.com/SigNoz/signoz/pkg/valuer"
 	"github.com/stretchr/testify/require"
 )
 
@@ -181,7 +182,7 @@ func TestStatementBuilder(t *testing.T) {
 	for _, c := range cases {
 		t.Run(c.name, func(t *testing.T) {
 
-			q, err := statementBuilder.Build(context.Background(), 1747947419000, 1747983448000, c.requestType, c.query, nil)
+			q, err := statementBuilder.Build(context.Background(), valuer.UUID{}, 1747947419000, 1747983448000, c.requestType, c.query, nil)
 
 			if c.expectedErr != nil {
 				require.Error(t, err)

pkg/telemetrymetrics/condition_builder.go

@@ -9,6 +9,7 @@ import (
 	"github.com/SigNoz/signoz/pkg/querybuilder"
 	qbtypes "github.com/SigNoz/signoz/pkg/types/querybuildertypes/querybuildertypesv5"
 	"github.com/SigNoz/signoz/pkg/types/telemetrytypes"
+	"github.com/SigNoz/signoz/pkg/valuer"
 
 	"github.com/huandu/go-sqlbuilder"
 )
@@ -143,6 +144,7 @@ func (c *conditionBuilder) conditionFor(
 
 func (c *conditionBuilder) ConditionFor(
 	ctx context.Context,
+	orgID valuer.UUID,
 	startNs uint64,
 	endNs uint64,
 	key *telemetrytypes.TelemetryFieldKey,

pkg/telemetrymetrics/condition_builder_test.go

@@ -6,6 +6,7 @@ import (
 
 	qbtypes "github.com/SigNoz/signoz/pkg/types/querybuildertypes/querybuildertypesv5"
 	"github.com/SigNoz/signoz/pkg/types/telemetrytypes"
+	"github.com/SigNoz/signoz/pkg/valuer"
 	"github.com/huandu/go-sqlbuilder"
 	"github.com/stretchr/testify/assert"
 	"github.com/stretchr/testify/require"
@@ -234,7 +235,7 @@ func TestConditionFor(t *testing.T) {
 	for _, tc := range testCases {
 		sb := sqlbuilder.NewSelectBuilder()
 		t.Run(tc.name, func(t *testing.T) {
-			cond, _, err := conditionBuilder.ConditionFor(ctx, 0, 0, &tc.key, []*telemetrytypes.TelemetryFieldKey{&tc.key}, tc.operator, tc.value, sb)
+			cond, _, err := conditionBuilder.ConditionFor(ctx, valuer.UUID{}, 0, 0, &tc.key, []*telemetrytypes.TelemetryFieldKey{&tc.key}, tc.operator, tc.value, sb)
 			sb.Where(cond...)
 
 			if tc.expectedError != nil {
@@ -289,7 +290,7 @@ func TestConditionForMultipleKeys(t *testing.T) {
 		t.Run(tc.name, func(t *testing.T) {
 			var err error
 			for _, key := range tc.keys {
-				cond, _, err := conditionBuilder.ConditionFor(ctx, 0, 0, &key, []*telemetrytypes.TelemetryFieldKey{&key}, tc.operator, tc.value, sb)
+				cond, _, err := conditionBuilder.ConditionFor(ctx, valuer.UUID{}, 0, 0, &key, []*telemetrytypes.TelemetryFieldKey{&key}, tc.operator, tc.value, sb)
 				sb.Where(cond...)
 				if err != nil {
 					t.Fatalf("Error getting condition for key %s: %v", key.Name, err)

pkg/telemetrymetrics/field_mapper.go

@@ -8,6 +8,7 @@ import (
 	schema "github.com/SigNoz/signoz-otel-collector/cmd/signozschemamigrator/schema_migrator"
 	qbtypes "github.com/SigNoz/signoz/pkg/types/querybuildertypes/querybuildertypesv5"
 	"github.com/SigNoz/signoz/pkg/types/telemetrytypes"
+	"github.com/SigNoz/signoz/pkg/valuer"
 	"github.com/huandu/go-sqlbuilder"
 )
 
@@ -37,6 +38,12 @@ var (
 
 type fieldMapper struct{}
 
+// CandidateKeys returns nil: metrics has no attribute-map fallback, so a context-missing
+// key stays unresolved and the caller errors.
+func (m *fieldMapper) CandidateKeys(_ context.Context, _ valuer.UUID, _ *telemetrytypes.TelemetryFieldKey, _ any, _ map[string][]*telemetrytypes.TelemetryFieldKey) []*telemetrytypes.TelemetryFieldKey {
+	return nil
+}
+
 func NewFieldMapper() qbtypes.FieldMapper {
 	return &fieldMapper{}
 }
@@ -92,6 +99,7 @@ func (m *fieldMapper) ColumnFor(ctx context.Context, tsStart, tsEnd uint64, key
 
 func (m *fieldMapper) ColumnExpressionFor(
 	ctx context.Context,
+	_ valuer.UUID,
 	startNs, endNs uint64,
 	field *telemetrytypes.TelemetryFieldKey,
 	keys map[string][]*telemetrytypes.TelemetryFieldKey,

pkg/telemetrymetrics/reduced_test.go

@@ -11,6 +11,7 @@ import (
 	qbtypes "github.com/SigNoz/signoz/pkg/types/querybuildertypes/querybuildertypesv5"
 	"github.com/SigNoz/signoz/pkg/types/telemetrytypes"
 	"github.com/SigNoz/signoz/pkg/types/telemetrytypes/telemetrytypestest"
+	"github.com/SigNoz/signoz/pkg/valuer"
 	"github.com/stretchr/testify/require"
 )
 
@@ -127,7 +128,7 @@ func TestReducedStatementBuilder(t *testing.T) {
 
 	for _, c := range cases {
 		t.Run(c.name, func(t *testing.T) {
-			got, err := sb.Build(context.Background(), start, end, qbtypes.RequestTypeTimeSeries, c.query, nil)
+			got, err := sb.Build(context.Background(), valuer.UUID{}, start, end, qbtypes.RequestTypeTimeSeries, c.query, nil)
 			require.NoError(t, err)
 			require.Equal(t, c.expected.Query, got.Query)
 			require.Equal(t, c.expected.Args, got.Args)
@@ -137,7 +138,7 @@ func TestReducedStatementBuilder(t *testing.T) {
 	t.Run("buffer_recent_window", func(t *testing.T) {
 		now := time.Now().UnixMilli()
 		q := reducedQuery("test.metric", metrictypes.GaugeType, metrictypes.Unspecified, metrictypes.TimeAggregationLatest, metrictypes.SpaceAggregationSum)
-		got, err := sb.Build(context.Background(), uint64(now-2*time.Hour.Milliseconds()), uint64(now), qbtypes.RequestTypeTimeSeries, q, nil)
+		got, err := sb.Build(context.Background(), valuer.UUID{}, uint64(now-2*time.Hour.Milliseconds()), uint64(now), qbtypes.RequestTypeTimeSeries, q, nil)
 		require.NoError(t, err)
 		require.Contains(t, got.Query, "signoz_metrics.distributed_samples_v4_buffer")
 		require.Contains(t, got.Query, "signoz_metrics.time_series_v4_buffer")
@@ -148,7 +149,7 @@ func TestReducedStatementBuilder(t *testing.T) {
 	t.Run("not_reduced", func(t *testing.T) {
 		q := reducedQuery("test.metric", metrictypes.GaugeType, metrictypes.Unspecified, metrictypes.TimeAggregationLatest, metrictypes.SpaceAggregationSum)
 		q.Aggregations[0].Reduced = false
-		got, err := sb.Build(context.Background(), start, end, qbtypes.RequestTypeTimeSeries, q, nil)
+		got, err := sb.Build(context.Background(), valuer.UUID{}, start, end, qbtypes.RequestTypeTimeSeries, q, nil)
 		require.NoError(t, err)
 		require.NotContains(t, got.Query, "UNION ALL")
 		require.NotContains(t, got.Query, "reduced")