feat: add configurable values for RED metric dimensions (#421)

Author: obs-gh-mattcotter

charts/agent/Chart.yaml

@@ -2,7 +2,7 @@ apiVersion: v2
 name: agent
 description: Chart to install K8s collection stack based on Observe Agent
 type: application
-version: 0.70.3
+version: 0.71.0
 appVersion: "2.8.1"
 dependencies:
   - name: opentelemetry-collector

charts/agent/README.md

@@ -1,6 +1,6 @@
 # agent
 
-![Version: 0.70.3](https://img.shields.io/badge/Version-0.70.3-informational?style=flat-square) ![Type: application](https://img.shields.io/badge/Type-application-informational?style=flat-square) ![AppVersion: 2.8.1](https://img.shields.io/badge/AppVersion-2.8.1-informational?style=flat-square)
+![Version: 0.71.0](https://img.shields.io/badge/Version-0.71.0-informational?style=flat-square) ![Type: application](https://img.shields.io/badge/Type-application-informational?style=flat-square) ![AppVersion: 2.8.1](https://img.shields.io/badge/AppVersion-2.8.1-informational?style=flat-square)
 
 Chart to install K8s collection stack based on Observe Agent
 
@@ -76,6 +76,8 @@ This service is a *single-instance deployment*. It's critical that this service
 | agent.selfMonitor.enabled | bool | `true` |  |
 | agent.selfMonitor.metrics.scrapeInterval | string | `"60s"` |  |
 | application.REDMetrics.enabled | bool | `false` | Whether to enable generating RED metrics from spans. See https://github.com/open-telemetry/opentelemetry-collector-contrib/tree/main/connector/spanmetricsconnector#overview |
+| application.REDMetrics.resourceDimensions | list | `["service.namespace","service.version","deployment.environment","k8s.pod.name","k8s.namespace.name"]` | List of resource attributes to include as dimensions for RED metrics. See https://github.com/open-telemetry/opentelemetry-collector-contrib/tree/main/connector/spanmetricsconnector#overview |
+| application.REDMetrics.spanDimensions | list | `["peer.db.name","peer.messaging.system","otel.status_description","observe.status_code"]` | List of span attributes to include as dimensions for RED metrics. See https://github.com/open-telemetry/opentelemetry-collector-contrib/tree/main/connector/spanmetricsconnector#overview |
 | application.prometheusScrape.enabled | bool | `false` |  |
 | application.prometheusScrape.independentDeployment | bool | `false` |  |
 | application.prometheusScrape.interval | string | `"60s"` |  |

charts/agent/templates/_config-connectors.tpl

@@ -1,6 +1,8 @@
 {{- define "config.connectors.spanmetrics" -}}
-{{/* Note: this must be kept in sync with the same variable in config.processors.RED_metrics */}}
-{{- $spanmetricsResourceAttributes := (list "service.namespace" "service.version" "deployment.environment" "k8s.pod.name" "k8s.namespace.name") -}}
+
+{{- $fullDimensions := (concat .Values.application.REDMetrics.resourceDimensions .Values.application.REDMetrics.spanDimensions) -}}
+{{/* service.name is added automatically by the connector and errors when specified in the config */}}
+{{- $fullDimensions = (without $fullDimensions "service.name" | uniq) -}}
 
 spanmetrics:
   aggregation_temporality: AGGREGATION_TEMPORALITY_DELTA
@@ -10,11 +12,7 @@ spanmetrics:
   dimensions:
     # This connector implicitly adds: service.name, span.name, span.kind, and status.code (which we rename to otel.status_code)
     # https://github.com/open-telemetry/opentelemetry-collector-contrib/blob/main/connector/spanmetricsconnector/connector.go#L528-L540
-    {{- range $tag := $spanmetricsResourceAttributes }}
+    {{- range $tag := $fullDimensions }}
     - name: {{ $tag }}
     {{- end }}
-    - name: peer.db.name
-    - name: peer.messaging.system
-    - name: otel.status_description
-    - name: observe.status_code
 {{- end -}}

charts/agent/templates/_config-processors.tpl

@@ -166,8 +166,6 @@ transform/add_span_status_code:
 {{- end -}}
 
 {{- define "config.processors.RED_metrics" -}}
-{{/* Note: this must be kept in sync with the same variable in config.connectors.spanmetrics */}}
-{{- $spanmetricsResourceAttributes := (list "service.namespace" "service.version" "deployment.environment" "k8s.pod.name" "k8s.namespace.name") -}}
 
 attributes/debug_source_span_metrics:
   actions:
@@ -220,16 +218,17 @@ filter/drop_span_kinds_other_than_server_and_consumer_and_peer_client:
       - span.kind == SPAN_KIND_PRODUCER
 {{- end }}
 
+{{- $resourceDims := (prepend .Values.application.REDMetrics.resourceDimensions "service.name" | uniq) }}
+
 # The spanmetrics connector puts all dimensions as attributes on the datapoint, and copies the resource attributes from an arbitrary span's resource. This cleans that up as well as handling any other renaming.
 transform/fix_red_metrics_resource_attributes:
   error_mode: ignore
   metric_statements:
     # Drop all resource attributes that aren't dimensions in the spanmetrics connector.
-    {{/* The service.name is implicit in the spanmetrics connector, so we can't include it in spanmetricsResourceAttributes and need to list it here. */}}
-    - keep_matching_keys(resource.attributes, "^(service.name|{{ join "|" $spanmetricsResourceAttributes }})")
+    - keep_matching_keys(resource.attributes, "^({{ join "|" $resourceDims }})")
 
     # Drop all datapoint attributes that are resource attributes in the spans.
-    - delete_matching_keys(datapoint.attributes, "^(service.name|{{ join "|" $spanmetricsResourceAttributes }})")
+    - delete_matching_keys(datapoint.attributes, "^({{ join "|" $resourceDims }})")
 
     # Rename status.code to response_status to be consistent with Trace Explorer and disambiguate from status_code (with an underscore).
     - set(datapoint.attributes["otel.status_code"], "OK") where datapoint.attributes["status.code"] == "STATUS_CODE_OK"

charts/agent/values.yaml

@@ -137,6 +137,19 @@ application:
   REDMetrics:
     # -- (bool) Whether to enable generating RED metrics from spans. See https://github.com/open-telemetry/opentelemetry-collector-contrib/tree/main/connector/spanmetricsconnector#overview
     enabled: false
+    # -- (list) List of resource attributes to include as dimensions for RED metrics. See https://github.com/open-telemetry/opentelemetry-collector-contrib/tree/main/connector/spanmetricsconnector#overview
+    resourceDimensions:
+      - service.namespace
+      - service.version
+      - deployment.environment
+      - k8s.pod.name
+      - k8s.namespace.name
+    # -- (list) List of span attributes to include as dimensions for RED metrics. See https://github.com/open-telemetry/opentelemetry-collector-contrib/tree/main/connector/spanmetricsconnector#overview
+    spanDimensions:
+      - peer.db.name
+      - peer.messaging.system
+      - otel.status_description
+      - observe.status_code
 
 gatewayDeployment:
   # -- (bool) Whether to enable the gateway deployment. The gateway is responsible for sampling and traces before sending them to Observe.