Inject dd-java-agent Using datakit-operator¶

Introduction¶

When connecting Java application APM in a Kubernetes environment, the dd-java-agent.jar package is required. To avoid modifying the application image, a common approach is to use initContainers in the deployment YAML and share the storage among containers in the same Pod to use dd-java-agent.jar. This approach results in the same initContainers section being present in every deployment file.

To extract these common parts and reduce workload further, you can use the open-source Admission Controller from Guance. datakit-operator injects dd-lib files and environment variables into specific Pods, currently supporting Java, Python, and JS.

Prerequisites¶

• You need to create a Guance account.
• A Kubernetes cluster.
• DataKit is already deployed via DaemonSet and the ddtrace collector is enabled.

Steps¶

1 Deploy datakit-operator¶

Download datakit-operator.yaml and deploy it to the Kubernetes cluster.

wget https://static.guance.com/datakit-operator/datakit-operator.yaml
kubectl apply -f datakit-operator.yaml

2 Deploy the Application¶

2.1 Write the Dockerfile¶

Write the Dockerfile for the application. The key point is to expose JAVA_OPTS so that the javaagent package can be injected externally.

FROM openjdk:8u292

RUN /bin/cp /usr/share/zoneinfo/Asia/Shanghai /etc/localtime
RUN echo 'Asia/Shanghai' >/etc/timezone

ENV jar service-demo-1.0-SNAPSHOT.jar

ENV workdir /data/app/
RUN mkdir -p ${workdir}
COPY ${jar} ${workdir}
WORKDIR ${workdir}
ENTRYPOINT ["sh", "-ec", "exec java ${JAVA_OPTS} -jar ${jar} ${PARAMS} "]

After writing, push the image to the registry. The following example directly pulls the image 172.16.0.246/df-demo/service-log-demo:v1.

2.2 Modify the Application YAML¶

Add an annotation to the application YAML.

      annotations:
        admission.datakit/java-lib.version: ""

2.3 Modify the Application YAML¶

Declare the JAVA_OPTS environment variable. The application can now access the automatically injected /datadog-lib/dd-java-agent.jar package. For startup parameters, refer to javaagent parameters. The service name set here is java-demo-service.

        - name: JAVA_OPTS
          value: |-
            -javaagent:/datadog-lib/dd-java-agent.jar -Ddd.service.name=java-demo-service -Ddd.tags=container_host:$(POD_NAME) -Ddd.env=dev -Ddd.agent.port=9529

apiVersion: v1
apiVersion: apps/v1
kind: Deployment
metadata:
  name: guance-deployment
spec:
  replicas: 1
  selector:
    matchLabels:
      app: guance-pod
  template:
    metadata:
      labels:
        app: guance-pod
      annotations:
        admission.datakit/java-lib.version: ""
    spec:
      containers:
      - name: guance-demo-container
        image: 172.16.0.246/df-demo/service-log-demo:v1
        env:
        - name: HOST_IP
          valueFrom:
            fieldRef:
              apiVersion: v1
              fieldPath: status.hostIP
        - name: HOST_NAME
          valueFrom:
            fieldRef:
              apiVersion: v1
              fieldPath: spec.nodeName
        - name: POD_NAME
          valueFrom:
            fieldRef:
              apiVersion: v1
              fieldPath: metadata.name
        #- name: DD_AGENT_HOST
        #  valueFrom:
        #    fieldRef:
        #      apiVersion: v1
        #      fieldPath: status.hostIP
        - name: JAVA_OPTS
          value: |-
            -javaagent:/datadog-lib/dd-java-agent.jar -Ddd.service.name=java-demo-service -Ddd.tags=container_host:$(POD_NAME) -Ddd.env=dev -Ddd.agent.port=9529
        ports:
        - containerPort: 8090
          protocol: TCP
        volumeMounts:
        - mountPath: /data/app/logs
          name: varlog
      restartPolicy: Always
      volumes:
      - name: varlog
        emptyDir: {}

2.4 Deploy the Application¶

kubectl apply -f java-demo.yaml

3 Trace Reporting¶

Get the Pod IP and access the application's endpoint to generate trace data.

Log in to Guance > APM, and search for the service java-demo-service to view the traces.