OTLP handler: enable metric name normalization (cortexproject#6136)

CHANGELOG.md

@@ -5,6 +5,7 @@
 * [CHANGE] Ingester: Remove `-querier.query-store-for-labels-enabled` flag. Querying long-term store for labels is always enabled. #5984
 * [CHANGE] Server: Instrument `cortex_request_duration_seconds` metric with native histogram. If `native-histograms` feature is enabled in monitoring Prometheus then the metric name needs to be updated in your dashboards. #6056
 * [CHANGE] Distributor/Ingester: Change `cortex_distributor_ingester_appends_total`, `cortex_distributor_ingester_append_failures_total`, `cortex_distributor_ingester_queries_total`, and `cortex_distributor_ingester_query_failures_total` metrics to use the ingester ID instead of its IP as the label value. #6078
+* [CHANGE] OTLP: Set `AddMetricSuffixes` to true to always enable metric name normalization. #6136
 * [FEATURE] Ingester/Distributor: Experimental: Enable native histogram ingestion via `-blocks-storage.tsdb.enable-native-histograms` flag. #5986 #6010 #6020
 * [FEATURE] Querier: Enable querying native histogram chunks. #5944 #6031
 * [FEATURE] Query Frontend: Support native histogram in query frontend response. #5996 #6043

pkg/util/push/otlp.go

@@ -38,7 +38,11 @@ func OTLPHandler(sourceIPs *middleware.SourceIPExtractor, push Func) http.Handle
 		}
 
 		promConverter := prometheusremotewrite.NewPrometheusConverter()
-		err = promConverter.FromMetrics(convertToMetricsAttributes(req.Metrics()), prometheusremotewrite.Settings{DisableTargetInfo: true})
+		setting := prometheusremotewrite.Settings{
+			AddMetricSuffixes: true,
+			DisableTargetInfo: true,
+		}
+		err = promConverter.FromMetrics(convertToMetricsAttributes(req.Metrics()), setting)
 		if err != nil {
 			level.Error(logger).Log("err", err.Error())
 			http.Error(w, err.Error(), http.StatusBadRequest)