add metrics endpoint for harness worker

.github/workflows/harness-image.yml

@@ -11,6 +11,11 @@ on:
   release:
     types: [published]
   workflow_dispatch:
+    inputs:
+      deploy:
+        description: "Dispatch downstream deploy after the image is built"
+        type: boolean
+        default: true
 
 concurrency:
   group: ${{ github.workflow }}-${{ github.ref }}
@@ -39,13 +44,21 @@ jobs:
           username: ${{ github.actor }}
           password: ${{ secrets.GITHUB_TOKEN }}
 
+      - name: Calculate branch tag
+        id: vars
+        shell: bash
+        run: |
+          BRANCH="${{ github.ref_name }}"
+          CLEANED_BRANCH_NAME=$(echo "$BRANCH" | tr '/' '-' | tr '[:upper:]' '[:lower:]')
+          echo "cleaned-branch-name=$CLEANED_BRANCH_NAME" >> "$GITHUB_OUTPUT"
+
       - name: Docker metadata
         id: meta
         uses: docker/metadata-action@v5
         with:
           images: ghcr.io/conductor-oss/java-sdk/harness-worker
           tags: |
-            type=raw,value=latest
+            type=raw,value=${{ steps.vars.outputs.cleaned-branch-name }}-latest,enable=${{ github.event_name != 'release' }}
             type=raw,value=${{ github.event.release.tag_name }},enable=${{ github.event_name == 'release' }}
 
       - name: Build and push
@@ -56,9 +69,17 @@ jobs:
           platforms: linux/amd64,linux/arm64
           push: true
           tags: ${{ steps.meta.outputs.tags }}
+          # Registry-backed BuildKit cache. Unchanged layers are reused across
+          # runs so rebuilding the same commit (or one with only minor diffs)
+          # is near-instant. The `:buildcache` tag lives alongside the image
+          # but only stores layer blobs, not a runnable image.
+          cache-from: type=registry,ref=ghcr.io/conductor-oss/java-sdk/harness-worker:buildcache
+          cache-to: type=registry,ref=ghcr.io/conductor-oss/java-sdk/harness-worker:buildcache,mode=max
 
   dispatch-deploy:
-    if: github.event_name == 'release'
+    if: |
+      github.event_name == 'release' ||
+      (github.event_name == 'workflow_dispatch' && inputs.deploy)
     needs: build-and-push
     runs-on: ubuntu-latest
     permissions:

harness/build.gradle

@@ -13,6 +13,7 @@ repositories {
 
 dependencies {
     implementation project(':conductor-client')
+    implementation project(':conductor-client-metrics')
     implementation project(':orkes-client')
     implementation "ch.qos.logback:logback-classic:1.5.6"
 }

harness/manifests/configmap-gcp.yaml

@@ -9,5 +9,5 @@ metadata:
   labels:
     app: java-sdk-harness-worker
 data:
-  CONDUCTOR_SERVER_URL: "https://certification-gcp.orkesconductor.com/api"
-  CONDUCTOR_AUTH_KEY: "e6c1ac61-286b-11f1-be01-c682b5750c3a"
+  CONDUCTOR_SERVER_URL: "https://certification-gcp.orkesconductor.io/api"
+  CONDUCTOR_AUTH_KEY: "25b681c1-34ec-11f1-b07a-9601c7a63373"

harness/manifests/deployment.yaml

@@ -17,7 +17,7 @@ spec:
     spec:
       containers:
       - name: harness
-        image: ghcr.io/conductor-oss/java-sdk/harness-worker:latest
+        image: ghcr.io/conductor-oss/java-sdk/harness-worker:main-latest
         imagePullPolicy: Always
         env:
         # === CONDUCTOR CONNECTION (from per-cloud ConfigMap) ===
@@ -52,6 +52,11 @@ spec:
         - name: HARNESS_POLL_INTERVAL_MS
           value: "100"
 
+        ports:
+        - name: metrics
+          containerPort: 9991
+          protocol: TCP
+
         resources:
           requests:
             memory: "256Mi"

harness/src/main/java/io/orkes/conductor/harness/HarnessMain.java

@@ -25,6 +25,7 @@
 import com.netflix.conductor.client.http.MetadataClient;
 import com.netflix.conductor.client.http.TaskClient;
 import com.netflix.conductor.client.http.WorkflowClient;
+import com.netflix.conductor.client.metrics.prometheus.PrometheusMetricsCollector;
 import com.netflix.conductor.client.worker.Worker;
 import com.netflix.conductor.common.metadata.tasks.TaskDef;
 import com.netflix.conductor.common.metadata.tasks.TaskType;
@@ -47,7 +48,7 @@ public class HarnessMain {
         {"java_worker_4", "deepcrawl", "5"},
     };
 
-    public static void main(String[] args) throws InterruptedException {
+    public static void main(String[] args) throws Exception {
         ConductorClient client = ApiClient.builder().useEnvVariables(true).readTimeout(10_000).connectTimeout(10_000)
                 .writeTimeout(10_000).build();
 
@@ -57,6 +58,11 @@ public static void main(String[] args) throws InterruptedException {
 
         registerMetadata(client);
 
+        PrometheusMetricsCollector metricsCollector = new PrometheusMetricsCollector();
+        int metricsPort = envInt("HARNESS_METRICS_PORT", 9991);
+        metricsCollector.startServer(metricsPort, "/metrics");
+        log.info("Prometheus metrics server started on port {}", metricsPort);
+
         List<Worker> workers = new ArrayList<>();
         for (String[] entry : SIMULATED_WORKERS) {
             workers.add(new SimulatedTaskWorker(entry[0], entry[1], Integer.parseInt(entry[2]), batchSize,
@@ -68,7 +74,10 @@ public static void main(String[] args) throws InterruptedException {
                 workers.stream().collect(Collectors.toMap(Worker::getTaskDefName, w -> batchSize));
 
         TaskRunnerConfigurer configurer =
-                new TaskRunnerConfigurer.Builder(taskClient, workers).withTaskThreadCount(threadCounts).build();
+                new TaskRunnerConfigurer.Builder(taskClient, workers)
+                        .withTaskThreadCount(threadCounts)
+                        .withMetricsCollector(metricsCollector)
+                        .build();
         configurer.init();
 
         WorkflowClient workflowClient = new WorkflowClient(client);