[POC][DO NOT MERGE] JMX Receiver POC

.gitlab-ci.yml

@@ -11,6 +11,7 @@ stages:
   - manaul-jobs
   - build
   - push
+  - kafka-poc-deploy
   - staging-deploy
   - prod-deploy
 
@@ -81,6 +82,22 @@ build-ci-image:
     # # For debugging
     # - aws sts get-caller-identity
     - bash $SCRIPT $NAMESPACE $VALUES
+kafka-poc-deploy:
+  stage: kafka-poc-deploy
+  tags: ["runner:docker", "size:large"]
+  image: $CI_IMAGE
+  dependencies:
+    - push-image-staging
+  rules:
+    - if: '$CI_COMMIT_REF_NAME =~ /-kafka-poc$/'
+  script:
+    - >-
+      TEMP_AWS_ACCESS_KEY_ID=$(aws ssm get-parameter --region us-east-1 --name ci.opentelemetry-collector-contrib.sand-eks-deploy-api-key --with-decryption --query Parameter.Value --out text)
+    - >-
+      TEMP_AWS_SECRET_ACCESS_KEY=$(aws ssm get-parameter --region us-east-1 --name ci.opentelemetry-collector-contrib.sand-eks-deploy-access-key --with-decryption --query Parameter.Value --out text)
+    - export AWS_ACCESS_KEY_ID=$TEMP_AWS_ACCESS_KEY_ID
+    - export AWS_SECRET_ACCESS_KEY=$TEMP_AWS_SECRET_ACCESS_KEY
+    - bash ./ci/scripts/kafka-poc.sh
 staging-deploy-demo-eks:
   !!merge <<: *staging-deploy
   variables:

ci/Dockerfile

@@ -12,7 +12,7 @@ RUN clean-apt install make \
 RUN apt-get update && \
 apt-get -y install default-jre-headless
 
-ADD https://github.com/open-telemetry/opentelemetry-java-contrib/releases/download/v1.27.0/opentelemetry-jmx-metrics.jar /opt/opentelemetry-jmx-metrics.jar
+ADD ./ci/opentelemetry-jmx-metrics-1.33.0-alpha-SNAPSHOT.jar /opt/opentelemetry-jmx-metrics.jar
 
 COPY --from=build /src/bin/otelcontribcol_* /otelcol-contrib
 USER 1001

ci/scripts/kafka-poc.sh

@@ -0,0 +1,49 @@
+#!/usr/bin/env bash
+
+# Copyright The OpenTelemetry Authors
+# SPDX-License-Identifier: Apache-2.0
+
+# This script is used to deploy collector on demo account cluster
+
+set -euo pipefail
+IFS=$'\n\t'
+set -x
+
+install_collector() {
+	release_name="opentelemetry-collector"
+
+	# if repo already exists, helm 3+ will skip
+	helm --debug repo add open-telemetry https://open-telemetry.github.io/opentelemetry-helm-charts
+	helm repo update open-telemetry
+	# --install will run `helm install` if not already present.
+	helm --debug upgrade "${release_name}" -n "kafka-poc" open-telemetry/opentelemetry-collector --install \
+		-f ./ci/values.yaml \
+		-f "./ci/values-kafka-poc.yaml" \
+		--set-string image.tag="otelcolcontrib-v$CI_COMMIT_SHORT_SHA" \
+		--set-string image.repository="601427279990.dkr.ecr.us-east-1.amazonaws.com/otel-collector-contrib"
+	helm list --all-namespaces
+
+	install_deployment
+}
+
+install_deployment() {
+	release_name_deployment="opentelemetry-collector-deployment"
+
+	# --install collector that fetches jmx metrics. The jmx receiver cannot be used in the daemonset deployment
+	# as this would lead to duplicate metrics.
+	helm --debug upgrade "${release_name_deployment}" -n "kafka-poc" open-telemetry/opentelemetry-collector --install \
+		-f ./ci/values-jmx.yaml \
+		--set-string image.tag="otelcolcontrib-v$CI_COMMIT_SHORT_SHA" \
+		--set-string image.repository="601427279990.dkr.ecr.us-east-1.amazonaws.com/otel-collector-contrib" \
+		--set nodeSelector.alpha\\.eksctl\\.io/nodegroup-name=kafka-poc-ng
+
+}
+
+###########################################################################################################
+clusterName="dd-otel"
+clusterArn="arn:aws:eks:us-east-1:601427279990:cluster/${clusterName}"
+
+aws eks --region us-east-1 update-kubeconfig --name "${clusterName}"
+kubectl config use-context "${clusterArn}"
+
+install_collector

ci/values-jmx.yaml

@@ -15,7 +15,7 @@ extraEnvs:
   - name: OTEL_EXPORTER_OTLP_ENDPOINT
     value: http://$(OTEL_COLLECTOR_NAME):4317
   - name: OTEL_RESOURCE_ATTRIBUTES
-    value: kafka_source=deploymentcol
+    value: kafka_source=otelcol
 config:
   receivers:
     jmx:

ci/values-kafka-poc.yaml

@@ -0,0 +1,6 @@
+nodeSelector:
+  alpha.eksctl.io/nodegroup-name: kafka-poc-ng
+clusterRole:
+  name: "opentelemetry-collector-kafka"
+  clusterRoleBinding:
+    name: opentelemetry-collector-kafka

ci/values.yaml

@@ -90,27 +90,46 @@ config:
     attributes:
       actions:
       - key: log.file.path
-        pattern: \/var\/log\/pods\/otel_opentelemetry-demo-(?P<service_name>.*?)-.*
+        pattern: \/var\/log\/pods\/${env:OTEL_K8S_NAMESPACE}_opentelemetry-demo-(?P<service_name>.*?)-.*
         action: extract
       - key: log.file.path
-        pattern: \/var\/log\/pods\/otel_opentelemetry-demo-(?P<source>.*?)-.*
+        pattern: \/var\/log\/pods\/${env:OTEL_K8S_NAMESPACE}_opentelemetry-demo-(?P<source>.*?)-.*
         action: extract
       - key: service
         from_attribute: service_name
         action: upsert
     attributes/kafkasrc:
       include:
-        match_type: regexp
+        match_type: strict
         attributes:
-          - {key: "source", value: 'kafka|frauddetectionservice|orderproducer'}
+          - key: "source"
+            value: 'kafka'
       actions:
       # this makes sure the source for logs that come from kafka|frauddetectionservice|orderproducer is
-      # set to kafka. The OOTB box kafka dashboard filters logs on source:kafka. This will still not work
-      # as it will be a log attribute and can only be searched by @source:kafka. We would need to support
-      # attributes_as_tags or find another solution for this to work.
-      - key: source
-        action: update
-        value: "kafka"
+      # set to kafka. The OOTB box kafka dashboard filters logs on source:kafka.
+      - key: ddtags
+        action: upsert
+        value: "source:kafka"
+    attributes/producer:
+      include:
+        match_type: strict
+        attributes:
+          - key: "source"
+            value: 'orderproducer'
+      actions:
+      - key: ddtags
+        action: upsert
+        value: "source:kafka"
+    attributes/consumer:
+      include:
+        match_type: strict
+        attributes:
+          - key: "source"
+            value: 'frauddetectionservice'
+      actions:
+      - key: ddtags
+        action: upsert
+        value: "source:kafka"
     memory_limiter:
       check_interval: 1s
       limit_mib: 500
@@ -192,5 +211,5 @@ config:
         processors: [resourcedetection, k8sattributes, batch]
         exporters: [datadog]
       logs:
-        processors: [memory_limiter, resourcedetection, k8sattributes, attributes, attributes/kafkasrc, batch]
+        processors: [memory_limiter, resourcedetection, k8sattributes, batch, attributes, attributes/kafkasrc, attributes/producer, attributes/consumer]
         exporters: [datadog]

cmd/otelcontribcol/go.mod

@@ -275,8 +275,9 @@ require (
 	github.com/DataDog/opentelemetry-mapping-go/pkg/inframetadata v0.8.1 // indirect
 	github.com/DataDog/opentelemetry-mapping-go/pkg/otlp/attributes v0.8.3 // indirect
 	github.com/DataDog/opentelemetry-mapping-go/pkg/otlp/logs v0.8.3 // indirect
-	github.com/DataDog/opentelemetry-mapping-go/pkg/otlp/metrics v0.8.1 // indirect
-	github.com/DataDog/opentelemetry-mapping-go/pkg/quantile v0.8.1 // indirect
+	// https://github.com/DataDog/opentelemetry-mapping-go/pull/217
+	github.com/DataDog/opentelemetry-mapping-go/pkg/otlp/metrics v0.8.4-0.20231214134201-f0bd121d625d // indirect
+	github.com/DataDog/opentelemetry-mapping-go/pkg/quantile v0.8.3 // indirect
 	github.com/DataDog/sketches-go v1.4.3 // indirect
 	github.com/DataDog/zstd v1.5.2 // indirect
 	github.com/GehirnInc/crypt v0.0.0-20200316065508-bb7000b8a962 // indirect

receiver/jmxreceiver/config.go

@@ -183,6 +183,9 @@ func (c *Config) validateJar(supportedJarDetails map[string]supportedJar, jar st
 		return fmt.Errorf("error hashing file: %w", err)
 	}
 
+	// Skip hash validation for POC.
+	return nil 
+
 	jarDetails, ok := supportedJarDetails[hash]
 	if !ok {
 		return errors.New("jar hash does not match known versions")