Merge branch 'main' into issue-5378

frontend/src/container/MetricsApplication/Tabs/Overview/ServiceOverview.tsx

@@ -1,6 +1,5 @@
 import { useMemo } from 'react';
 import { useParams } from 'react-router-dom';
-import { Skeleton } from 'antd';
 import { ENTITY_VERSION_V4 } from 'constants/app';
 import { FeatureKeys } from 'constants/features';
 import { PANEL_TYPES } from 'constants/queryBuilder';
@@ -124,24 +123,14 @@ function ServiceOverview({
 			/>
 			<Card data-testid="service_latency">
 				<GraphContainer>
-					{topLevelOperationsIsLoading && (
-						<Skeleton
-							style={{
-								height: '100%',
-								padding: '16px',
-							}}
-						/>
-					)}
-					{!topLevelOperationsIsLoading && (
-						<Graph
-							onDragSelect={onDragSelect}
-							widget={latencyWidget}
-							onClickHandler={handleGraphClick('Service')}
-							isQueryEnabled={isQueryEnabled}
-							version={ENTITY_VERSION_V4}
-							enableDrillDown={SERVICE_DETAIL_DRILLDOWN_ENABLED}
-						/>
-					)}
+					<Graph
+						onDragSelect={onDragSelect}
+						widget={latencyWidget}
+						onClickHandler={handleGraphClick('Service')}
+						isQueryEnabled={isQueryEnabled}
+						version={ENTITY_VERSION_V4}
+						enableDrillDown={SERVICE_DETAIL_DRILLDOWN_ENABLED}
+					/>
 				</GraphContainer>
 			</Card>
 		</>

frontend/src/container/MetricsApplication/Tabs/Overview/TopLevelOperations.tsx

@@ -1,4 +1,3 @@
-import { Skeleton } from 'antd';
 import { Typography } from '@signozhq/ui/typography';
 import axios from 'axios';
 import { SOMETHING_WENT_WRONG } from 'constants/api';
@@ -29,24 +28,14 @@ function TopLevelOperation({
 				</Typography>
 			) : (
 				<GraphContainer>
-					{topLevelOperationsIsLoading && (
-						<Skeleton
-							style={{
-								height: '100%',
-								padding: '16px',
-							}}
-						/>
-					)}
-					{!topLevelOperationsIsLoading && (
-						<Graph
-							widget={widget}
-							onClickHandler={handleGraphClick(opName)}
-							onDragSelect={onDragSelect}
-							isQueryEnabled={!topLevelOperationsIsLoading}
-							version={ENTITY_VERSION_V4}
-							enableDrillDown={SERVICE_DETAIL_DRILLDOWN_ENABLED}
-						/>
-					)}
+					<Graph
+						widget={widget}
+						onClickHandler={handleGraphClick(opName)}
+						onDragSelect={onDragSelect}
+						isQueryEnabled={!topLevelOperationsIsLoading}
+						version={ENTITY_VERSION_V4}
+						enableDrillDown={SERVICE_DETAIL_DRILLDOWN_ENABLED}
+					/>
 				</GraphContainer>
 			)}
 		</Card>

pkg/modules/cloudintegration/config.go

@@ -22,7 +22,7 @@ func newConfig() factory.Config {
 		Agent: AgentConfig{
 			// we will maintain the latest version of cloud integration agent from here,
 			// till we automate it externally or figure out a way to validate it.
-			Version: "v0.0.12",
+			Version: "v0.0.13",
 		},
 	}
 }

pkg/querier/querier.go

@@ -119,11 +119,7 @@ func (q *querier) QueryRange(ctx context.Context, orgID valuer.UUID, req *qbtype
 	queries := make(map[string]qbtypes.Query)
 	steps := make(map[string]qbtypes.Step)
 
-	// Resolve metric metadata once per request: patches each metric-aggregation
-	// query's spec in place, returns the queries whose every aggregation was
-	// missing (used for preseeded empty results), and any dormant-metric
-	// warning string. NotFound errors for never-seen metrics are propagated.
-	missingMetricQueries, dormantMetricsWarningMsg, err := q.resolveMetricMetadata(ctx, req.CompositeQuery.Queries, req.Start, req.End)
+	missingMetricQueries, metricWarnings, err := q.resolveMetricMetadata(ctx, req.CompositeQuery.Queries, req.Start, req.End)
 	if err != nil {
 		return nil, err
 	}
@@ -240,13 +236,15 @@ func (q *querier) QueryRange(ctx context.Context, orgID valuer.UUID, req *qbtype
 				}
 			}
 		}
-		if dormantMetricsWarningMsg != "" {
+		if len(metricWarnings) > 0 {
 			if qbResp.Warning == nil {
 				qbResp.Warning = &qbtypes.QueryWarnData{}
 			}
-			qbResp.Warning.Warnings = append(qbResp.Warning.Warnings, qbtypes.QueryWarnDataAdditional{
-				Message: dormantMetricsWarningMsg,
-			})
+			for _, w := range metricWarnings {
+				qbResp.Warning.Warnings = append(qbResp.Warning.Warnings, qbtypes.QueryWarnDataAdditional{
+					Message: w,
+				})
+			}
 		}
 	}
 	return qbResp, qbErr
@@ -302,12 +300,11 @@ func (q *querier) populateQBEvent(event *qbtypes.QBEvent, queries []qbtypes.Quer
 //   - missingMetricQueries: names of queries whose every aggregation was
 //     missing. Used downstream to preseed empty result placeholders so the
 //     response still has an entry per requested query name.
-//   - dormantWarning: a human-readable warning describing metrics that exist in
-//     the store but produced no data within the query window. Empty when no
-//     such metrics are present.
-//   - err: NotFound when one or more referenced metrics have never been seen,
-//     or Internal when a metadata fetch fails.
-func (q *querier) resolveMetricMetadata(ctx context.Context, queries []qbtypes.QueryEnvelope, start, end uint64) (missingMetricQueries []string, dormantWarning string, err error) {
+//   - metricWarnings: human-readable warnings for metrics that could not be
+//     resolved: never-seen metrics and dormant metrics (seen but no data in
+//     the query window).
+//   - err: Internal when a metadata fetch fails.
+func (q *querier) resolveMetricMetadata(ctx context.Context, queries []qbtypes.QueryEnvelope, start, end uint64) (missingMetricQueries []string, metricWarnings []string, err error) {
 	metricNames := make([]string, 0)
 	for idx := range queries {
 		if queries[idx].Type != qbtypes.QueryTypeBuilder {
@@ -325,13 +322,13 @@ func (q *querier) resolveMetricMetadata(ctx context.Context, queries []qbtypes.Q
 	}
 
 	if len(metricNames) == 0 {
-		return nil, "", nil
+		return nil, nil, nil
 	}
 
 	metricTemporality, metricTypes, err := q.metadataStore.FetchTemporalityAndTypeMulti(ctx, start, end, metricNames...)
 	if err != nil {
 		q.logger.WarnContext(ctx, "failed to fetch metric temporality", errors.Attr(err), slog.Any("metrics", metricNames))
-		return nil, "", errors.NewInternalf(errors.CodeInternal, "failed to fetch metrics temporality")
+		return nil, nil, errors.NewInternalf(errors.CodeInternal, "failed to fetch metrics temporality")
 	}
 	q.logger.DebugContext(ctx, "fetched metric temporalities and types", slog.Any("metric_temporality", metricTemporality), slog.Any("metric_types", metricTypes))
 
@@ -363,7 +360,7 @@ func (q *querier) resolveMetricMetadata(ctx context.Context, queries []qbtypes.Q
 			}
 			// Type is resolved now; validate aggregation compatibility against it.
 			if err := spec.Aggregations[i].ValidateForType(); err != nil {
-				return nil, "", err
+				return nil, nil, err
 			}
 			presentAggregations = append(presentAggregations, spec.Aggregations[i])
 		}
@@ -376,7 +373,7 @@ func (q *querier) resolveMetricMetadata(ctx context.Context, queries []qbtypes.Q
 	}
 
 	if len(missingMetrics) == 0 {
-		return missingMetricQueries, "", nil
+		return missingMetricQueries, nil, nil
 	}
 
 	isInternalMetric := func(n string) bool { return strings.HasPrefix(n, "signoz.") || strings.HasPrefix(n, "signoz_") }
@@ -387,29 +384,33 @@ func (q *querier) resolveMetricMetadata(ctx context.Context, queries []qbtypes.Q
 		}
 	}
 	if len(externalMissingMetrics) == 0 {
-		// this means all missing metrics are internal, and since internal metrics
-		// aren't user-controlled, skip errors/warnings for them since users can't act on them
-		return missingMetricQueries, "", nil
+		return missingMetricQueries, nil, nil
 	}
 
-	// Classify each missing metric: never-seen → NotFound error; seen-but-no-
-	// data-in-window → dormant warning.
+	// Classify each missing metric: never-seen -> warning with empty result;
+	// seen-but-no-data-in-window -> dormant warning.
 	lastSeenInfo, _ := q.metadataStore.FetchLastSeenInfoMulti(ctx, externalMissingMetrics...)
-	nonExistentMetrics := []string{}
+	var nonExistentMetrics []string
+	var dormantMetrics []string
 	for _, name := range externalMissingMetrics {
 		if ts, ok := lastSeenInfo[name]; ok && ts > 0 {
+			dormantMetrics = append(dormantMetrics, name)
 			continue
 		}
 		nonExistentMetrics = append(nonExistentMetrics, name)
 	}
+
+	var warnings []string
+
+	// Never-seen metrics: the query already gets a preseeded empty result
+	// via the aggregation-dropping path above; we just attach a warning.
 	if len(nonExistentMetrics) == 1 {
-		return nil, "", errors.NewNotFoundf(errors.CodeNotFound, "could not find the metric %s", nonExistentMetrics[0])
-	}
-	if len(nonExistentMetrics) > 1 {
-		return nil, "", errors.NewNotFoundf(errors.CodeNotFound, "the following metrics were not found: %s", strings.Join(nonExistentMetrics, ", "))
+		warnings = append(warnings, fmt.Sprintf("metric %s has never been received. Check the metric name and instrumentation", nonExistentMetrics[0]))
+	} else if len(nonExistentMetrics) > 1 {
+		warnings = append(warnings, fmt.Sprintf("the following metrics have never been received. Check the metric names and instrumentation: %s", strings.Join(nonExistentMetrics, ", ")))
 	}
 
-	// All missing metrics are dormant — assemble the warning string.
+	// Dormant metrics: seen before but no data in the query window.
 	lastSeenStr := func(name string) string {
 		if ts, ok := lastSeenInfo[name]; ok && ts > 0 {
 			ago := humanize.RelTime(time.UnixMilli(ts), time.Now(), "ago", "from now")
@@ -417,16 +418,16 @@ func (q *querier) resolveMetricMetadata(ctx context.Context, queries []qbtypes.Q
 		}
 		return name
 	}
-	if len(externalMissingMetrics) == 1 {
-		dormantWarning = fmt.Sprintf("no data found for the metric %s in the query time range", lastSeenStr(missingMetrics[0]))
-	} else {
-		parts := make([]string, len(externalMissingMetrics))
-		for i, m := range externalMissingMetrics {
+	if len(dormantMetrics) == 1 {
+		warnings = append(warnings, fmt.Sprintf("no data found for the metric %s in the query time range", lastSeenStr(dormantMetrics[0])))
+	} else if len(dormantMetrics) > 1 {
+		parts := make([]string, len(dormantMetrics))
+		for i, m := range dormantMetrics {
 			parts[i] = lastSeenStr(m)
 		}
-		dormantWarning = fmt.Sprintf("no data found for the following metrics in the query time range: %s", strings.Join(parts, ", "))
+		warnings = append(warnings, fmt.Sprintf("no data found for the following metrics in the query time range: %s", strings.Join(parts, ", ")))
 	}
-	return missingMetricQueries, dormantWarning, nil
+	return missingMetricQueries, warnings, nil
 }
 
 func (q *querier) QueryRawStream(ctx context.Context, orgID valuer.UUID, req *qbtypes.QueryRangeRequest, client *qbtypes.RawStream) {

pkg/querier/querier_test.go

@@ -37,7 +37,7 @@ func (m *mockMetricStmtBuilder) Build(_ context.Context, _, _ uint64, _ qbtypes.
 
 func TestQueryRange_MetricTypeMissing(t *testing.T) {
 	// When a metric has UnspecifiedType and is not found in the metadata store,
-	// the querier should return a not-found error, even if the request provides a temporality
+	// the querier should return an empty result with a warning instead of an error.
 	providerSettings := instrumentationtest.New().ToProviderSettings()
 	metadataStore := telemetrytypestest.NewMockMetadataStore()
 
@@ -80,9 +80,14 @@ func TestQueryRange_MetricTypeMissing(t *testing.T) {
 		},
 	}
 
-	_, err := q.QueryRange(context.Background(), valuer.GenerateUUID(), req)
-	require.Error(t, err)
-	assert.Contains(t, err.Error(), "could not find the metric unknown_metric")
+	resp, err := q.QueryRange(context.Background(), valuer.GenerateUUID(), req)
+	require.NoError(t, err)
+	require.NotNil(t, resp)
+	require.NotNil(t, resp.Warning)
+
+	require.Len(t, resp.Warning.Warnings, 1)
+	assert.Contains(t, resp.Warning.Warnings[0].Message, "unknown_metric")
+	assert.Contains(t, resp.Warning.Warnings[0].Message, "has never been received")
 }
 
 func TestQueryRange_MetricTypeFromStore(t *testing.T) {

pkg/telemetrymetrics/statement_builder.go

@@ -101,9 +101,29 @@ func (b *MetricQueryStatementBuilder) Build(
 		return nil, err
 	}
 
+	var pairFallbackWarnings []string
+	for _, sel := range keySelectors {
+		if _, ok := keys[sel.Name]; !ok {
+			keys[sel.Name] = []*telemetrytypes.TelemetryFieldKey{{
+				Name:          sel.Name,
+				FieldContext:  telemetrytypes.FieldContextAttribute,
+				FieldDataType: telemetrytypes.FieldDataTypeString,
+				Signal:        telemetrytypes.SignalMetrics,
+			}}
+			pairFallbackWarnings = append(pairFallbackWarnings,
+				fmt.Sprintf("key `%s` not found on metric %s", sel.Name, query.Aggregations[0].MetricName),
+			)
+		}
+	}
+
 	start, end = querybuilder.AdjustedMetricTimeRange(start, end, uint64(query.StepInterval.Seconds()), query)
 
-	return b.buildPipelineStatement(ctx, start, end, query, keys, variables)
+	stmt, err := b.buildPipelineStatement(ctx, start, end, query, keys, variables)
+	if err != nil {
+		return nil, err
+	}
+	stmt.Warnings = append(stmt.Warnings, pairFallbackWarnings...)
+	return stmt, nil
 }
 
 func (b *MetricQueryStatementBuilder) buildPipelineStatement(

pkg/telemetrymetrics/stmt_builder_test.go

@@ -217,6 +217,39 @@ func TestStatementBuilder(t *testing.T) {
 			},
 			expectedErr: nil,
 		},
+		{
+			name:        "test_missing_key_falls_back_to_labels",
+			requestType: qbtypes.RequestTypeTimeSeries,
+			query: qbtypes.QueryBuilderQuery[qbtypes.MetricAggregation]{
+				Signal:       telemetrytypes.SignalMetrics,
+				StepInterval: qbtypes.Step{Duration: 30 * time.Second},
+				Aggregations: []qbtypes.MetricAggregation{
+					{
+						MetricName:       "signoz_calls_total",
+						Type:             metrictypes.SumType,
+						Temporality:      metrictypes.Cumulative,
+						TimeAggregation:  metrictypes.TimeAggregationRate,
+						SpaceAggregation: metrictypes.SpaceAggregationSum,
+					},
+				},
+				Filter: &qbtypes.Filter{
+					Expression: "k8s.statefulset.name = 'my-statefulset'",
+				},
+				GroupBy: []qbtypes.GroupByKey{
+					{
+						TelemetryFieldKey: telemetrytypes.TelemetryFieldKey{
+							Name: "k8s.statefulset.name",
+						},
+					},
+				},
+			},
+			expected: qbtypes.Statement{
+				Query: "WITH __temporal_aggregation_cte AS (SELECT ts, `k8s.statefulset.name`, multiIf(row_number() OVER rate_window = 1, nan, (per_series_value - lagInFrame(per_series_value, 1) OVER rate_window) < 0, per_series_value / (ts - lagInFrame(ts, 1) OVER rate_window), (per_series_value - lagInFrame(per_series_value, 1) OVER rate_window) / (ts - lagInFrame(ts, 1) OVER rate_window)) AS per_series_value FROM (SELECT fingerprint, toStartOfInterval(toDateTime(intDiv(unix_milli, 1000)), toIntervalSecond(30)) AS ts, `k8s.statefulset.name`, max(value) AS per_series_value FROM signoz_metrics.distributed_samples_v4 AS points INNER JOIN (SELECT fingerprint, JSONExtractString(labels, 'k8s.statefulset.name') AS `k8s.statefulset.name` FROM signoz_metrics.time_series_v4_6hrs WHERE metric_name IN (?) AND unix_milli >= ? AND unix_milli <= ? AND LOWER(temporality) LIKE LOWER(?) AND __normalized = ? AND JSONExtractString(labels, 'k8s.statefulset.name') = ? GROUP BY fingerprint, `k8s.statefulset.name`) AS filtered_time_series ON points.fingerprint = filtered_time_series.fingerprint WHERE metric_name IN (?) AND unix_milli >= ? AND unix_milli < ? GROUP BY fingerprint, ts, `k8s.statefulset.name` ORDER BY fingerprint, ts) WINDOW rate_window AS (PARTITION BY fingerprint ORDER BY fingerprint, ts)), __spatial_aggregation_cte AS (SELECT ts, `k8s.statefulset.name`, sum(per_series_value) AS value FROM __temporal_aggregation_cte WHERE isNaN(per_series_value) = ? GROUP BY ts, `k8s.statefulset.name`) SELECT * FROM __spatial_aggregation_cte ORDER BY `k8s.statefulset.name`, ts",
+				Args:     []any{"signoz_calls_total", uint64(1747936800000), uint64(1747983420000), "cumulative", false, "my-statefulset", "signoz_calls_total", uint64(1747947360000), uint64(1747983420000), 0},
+				Warnings: []string{"key `k8s.statefulset.name` not found on metric signoz_calls_total"},
+			},
+			expectedErr: nil,
+		},
 	}
 
 	fm := NewFieldMapper()

tests/integration/tests/querier/03_metrics.py

@@ -614,7 +614,7 @@ def test_histogram_p90_returns_warning_outside_data_window(
     assert warnings[0]["message"].startswith(f"no data found for the metric {metric_name}")
 
 
-def test_non_existent_metrics_returns_404(
+def test_non_existent_metrics_returns_warning(
     signoz: types.SigNoz,
     create_user_admin: None,  # pylint: disable=unused-argument
     get_token: Callable[[str, str], str],
@@ -635,9 +635,11 @@ def test_non_existent_metrics_returns_404(
 
     start_2h = int((now - timedelta(hours=2)).timestamp() * 1000)
     response = make_query_request(signoz, token, start_2h, end_ms, [query])
-    assert response.status_code == HTTPStatus.NOT_FOUND
+    assert response.status_code == HTTPStatus.OK
 
-    assert get_error_message(response.json()) == "could not find the metric whatevergoennnsgoeshere"
+    data = response.json()
+    warnings = get_all_warnings(data)
+    assert any("whatevergoennnsgoeshere" in w["message"] and "has never been received" in w["message"] for w in warnings), f"expected never-seen metric warning, got: {warnings}"
 
 
 def test_non_existent_internal_metrics_returns_no_warning(