ci: build to GHCR and deploy to do-prod

.github/workflows/deploy-to-dev.yml

@@ -1,4 +1,7 @@
-name: Deploy Survey to Dev (GitOps)
+# Post AWS->DO migration: main builds/pushes to GHCR and deploys straight to
+# do-prod (the only live cluster — EKS dev/prod are retired). Filename kept to
+# preserve any required-status-check config; see the job below for the real target.
+name: Build & Deploy Survey to Prod (GitOps)
 
 on:
   push:
@@ -29,7 +32,11 @@ jobs:
           cache-dependency-path: web/package-lock.json
 
       - name: Install templ
-        run: go install github.com/a-h/templ/cmd/templ@latest
+        # Pin the CLI to the go.mod runtime version (github.com/a-h/templ) so
+        # `templ generate` emits code the pinned runtime can compile. `@latest`
+        # drifts ahead of the runtime (e.g. templ.ResolveAttributeValue) and
+        # breaks the build. Bump this in lockstep with go.mod.
+        run: go install github.com/a-h/templ/cmd/templ@v0.3.960
 
       - name: Install dependencies
         run: go mod download
@@ -52,46 +59,42 @@ jobs:
   build-and-deploy:
     runs-on: ubuntu-latest
     if: github.event_name == 'push' && github.ref == 'refs/heads/main'
+    permissions:
+      contents: read
+      packages: write   # push to ghcr.io/openmeet-team/survey via GITHUB_TOKEN
     outputs:
       image_tag: ${{ steps.set-outputs.outputs.image_tag }}
 
     steps:
       - name: Checkout repository
         uses: actions/checkout@v4
 
-      - name: Configure AWS credentials
-        uses: aws-actions/configure-aws-credentials@v1
+      # Auth to GHCR. GITHUB_TOKEN works once the package is linked to this repo
+      # (Package settings -> Manage Actions access -> add survey = Write).
+      # To use a PAT instead: set password to ${{ secrets.GHCR_PAT }} and drop the
+      # packages: write permission above.
+      - name: Log in to GHCR
+        uses: docker/login-action@v3
         with:
-          aws-access-key-id: ${{ secrets.AWS_ACCESS_KEY_ID }}
-          aws-secret-access-key: ${{ secrets.AWS_SECRET_ACCESS_KEY }}
-          aws-region: ${{ vars.AWS_REGION || 'us-east-1' }}
-
-      - name: Login to Amazon ECR
-        id: login-ecr
-        uses: aws-actions/amazon-ecr-login@v2
-        with:
-          mask-password: true
+          registry: ghcr.io
+          username: ${{ github.actor }}
+          password: ${{ secrets.GITHUB_TOKEN }}
 
       - name: Set image tag
         id: set-outputs
         run: |
           echo "IMAGE_TAG=${{ github.sha }}" >> $GITHUB_ENV
           echo "image_tag=${{ github.sha }}" >> $GITHUB_OUTPUT
 
-      - name: Build, tag, and push image to Amazon ECR
+      - name: Build and push image to GHCR
         env:
-          ECR_REGISTRY: ${{ steps.login-ecr.outputs.registry }}
-          ECR_REPOSITORY: openmeet-ecr/survey
-          GIT_REVISION: ${{ github.sha }}
-          GIT_BRANCH: ${{ github.ref_name }}
+          IMAGE: ghcr.io/openmeet-team/survey
         run: |
-          echo "Building image with tag: $ECR_REGISTRY/$ECR_REPOSITORY:$IMAGE_TAG"
-          docker build -t $ECR_REGISTRY/$ECR_REPOSITORY:$IMAGE_TAG .
-          docker push $ECR_REGISTRY/$ECR_REPOSITORY:$IMAGE_TAG
-
-          # Tag as latest for main branch
-          docker tag $ECR_REGISTRY/$ECR_REPOSITORY:$IMAGE_TAG $ECR_REGISTRY/$ECR_REPOSITORY:latest
-          docker push $ECR_REGISTRY/$ECR_REPOSITORY:latest
+          set -euo pipefail
+          echo "Building $IMAGE:$IMAGE_TAG"
+          docker build -t "$IMAGE:$IMAGE_TAG" -t "$IMAGE:latest" .
+          docker push "$IMAGE:$IMAGE_TAG"
+          docker push "$IMAGE:latest"
 
   gitops-update:
     needs: [build-and-deploy]
@@ -100,9 +103,8 @@ jobs:
     with:
       service_name: survey
       image_tag: ${{ needs.build-and-deploy.outputs.image_tag }}
-      environment: dev
-      ecr_registry: 433321780850.dkr.ecr.us-east-1.amazonaws.com
-      ecr_repository: openmeet-ecr/survey
-      commit_message: "GitOps: Update Survey to ${{ needs.build-and-deploy.outputs.image_tag }} from main branch"
+      image_name: ghcr.io/openmeet-team/survey
+      environment: do-prod
+      commit_message: "GitOps: deploy survey ${{ needs.build-and-deploy.outputs.image_tag }} to do-prod (main)"
     secrets:
       GH_PAT: ${{ secrets.GH_PAT_INFRASTRUCTURE }}

.github/workflows/update-image-gitops.yml

@@ -4,33 +4,30 @@ on:
   workflow_call:
     inputs:
       service_name:
-        description: 'Service name (api, platform, survey, etc.)'
+        description: 'Service name (api, platform, biz, ...) — used for log/commit context'
         required: true
         type: string
       image_tag:
-        description: 'Docker image tag to deploy'
+        description: 'Image tag (git SHA) to deploy'
         required: true
         type: string
-      environment:
-        description: 'Target environment (dev, prod, pr-123, etc.)'
-        required: true
-        type: string
-      ecr_registry:
-        description: 'ECR registry URL'
+      image_name:
+        description: 'Full GHCR image name to pin (matched against images[].newName), e.g. ghcr.io/openmeet-team/openmeet-api'
         required: true
         type: string
-      ecr_repository:
-        description: 'ECR repository name'
-        required: true
+      environment:
+        description: 'Target overlay under k8s/environments/ (do-prod is the only live cluster post-migration)'
+        required: false
         type: string
+        default: do-prod
       commit_message:
         description: 'Git commit message'
         required: false
         type: string
         default: 'Update image tag'
     secrets:
       GH_PAT:
-        description: 'GitHub PAT with repo access'
+        description: 'GitHub PAT with write access to the openmeet-infrastructure repo'
         required: true
 
 jobs:
@@ -44,67 +41,76 @@ jobs:
           token: ${{ secrets.GH_PAT }}
           ref: main
 
-      - name: Update kustomization.yaml with new image tag
+      - name: Pin ${{ inputs.service_name }} image to ${{ inputs.image_tag }}
+        env:
+          IMAGE_NAME: ${{ inputs.image_name }}
+          IMAGE_TAG: ${{ inputs.image_tag }}
+          ENVIRONMENT: ${{ inputs.environment }}
         run: |
-          KUSTOMIZATION_PATH="k8s/environments/${{ inputs.environment }}/kustomization.yaml"
-          IMAGE_FULL="${{ inputs.ecr_registry }}/${{ inputs.ecr_repository }}:${{ inputs.image_tag }}"
-
-          echo "Updating $KUSTOMIZATION_PATH"
-          echo "Service: ${{ inputs.service_name }}"
-          echo "New image: $IMAGE_FULL"
-
-          # Update the image tag in the kustomization file
-          # This uses sed to find the line with the service image and update it
-          sed -i "s|image: .*${{ inputs.ecr_repository }}:.*|image: $IMAGE_FULL|" "$KUSTOMIZATION_PATH"
+          set -euo pipefail
+          KUSTOMIZATION_PATH="k8s/environments/${ENVIRONMENT}/kustomization.yaml"
+          echo "Pinning $IMAGE_NAME -> $IMAGE_TAG in $KUSTOMIZATION_PATH"
+
+          # Single source of truth: the overlay's `images:` transformer. Set newTag
+          # on EVERY entry whose newName is this service's GHCR image. Keying on
+          # newName (not the ECR lookup key) updates all references at once —
+          # e.g. survey has two entries (bare `survey` + the ECR name) that share
+          # one newName. yq (mikefarah) is preinstalled on ubuntu-latest runners.
+          yq -i '(.images[] | select(.newName == strenv(IMAGE_NAME))).newTag = strenv(IMAGE_TAG)' "$KUSTOMIZATION_PATH"
+
+          # Guard: fail loudly if nothing matched (typo in image_name, or the
+          # overlay has no images[] entry for this service) — never push a no-op
+          # that silently fails to deploy.
+          matched="$(yq '[.images[] | select(.newName == strenv(IMAGE_NAME) and .newTag == strenv(IMAGE_TAG))] | length' "$KUSTOMIZATION_PATH")"
+          if [ "$matched" -eq 0 ]; then
+            echo "::error::No images[] entry with newName=$IMAGE_NAME in $KUSTOMIZATION_PATH — nothing to deploy."
+            exit 1
+          fi
+          echo "Updated $matched image reference(s)."
 
-          # Show the diff
           git diff "$KUSTOMIZATION_PATH"
 
       - name: Commit and push changes
+        env:
+          ENVIRONMENT: ${{ inputs.environment }}
+          COMMIT_MESSAGE: ${{ inputs.commit_message }}
         run: |
+          set -euo pipefail
           git config --global user.email "github-actions@github.com"
           git config --global user.name "GitHub Actions Bot"
 
-          KUSTOMIZATION_PATH="k8s/environments/${{ inputs.environment }}/kustomization.yaml"
-
+          KUSTOMIZATION_PATH="k8s/environments/${ENVIRONMENT}/kustomization.yaml"
           git add "$KUSTOMIZATION_PATH"
 
-          # Check if there are changes to commit
           if git diff --staged --quiet; then
-            echo "No changes to commit - image tag may already be up to date"
+            echo "No changes to commit — image tag already up to date."
             exit 0
           fi
 
-          git commit -m "${{ inputs.commit_message }}"
+          git commit -m "$COMMIT_MESSAGE"
 
-          # Push with retry logic to handle concurrent updates
+          # Push with retry to handle concurrent gitops updates from sibling repos.
           MAX_RETRIES=5
           RETRY_COUNT=0
-
           while [ $RETRY_COUNT -lt $MAX_RETRIES ]; do
             echo "Attempting to push (attempt $((RETRY_COUNT + 1))/$MAX_RETRIES)..."
-
             if git push origin main; then
-              echo "Successfully pushed changes"
+              echo "✅ Successfully pushed changes"
               exit 0
-            else
-              echo "Push failed, pulling and retrying..."
-              RETRY_COUNT=$((RETRY_COUNT + 1))
-
-              if [ $RETRY_COUNT -lt $MAX_RETRIES ]; then
-                # Pull with rebase to get latest changes
-                git pull --rebase origin main
-                sleep $((RETRY_COUNT * 2))  # Exponential backoff
-              fi
+            fi
+            echo "⚠️  Push failed, pulling and retrying..."
+            RETRY_COUNT=$((RETRY_COUNT + 1))
+            if [ $RETRY_COUNT -lt $MAX_RETRIES ]; then
+              git pull --rebase origin main
+              sleep $((RETRY_COUNT * 2))
             fi
           done
 
-          echo "Failed to push after $MAX_RETRIES attempts"
+          echo "❌ Failed to push after $MAX_RETRIES attempts"
           exit 1
 
-      - name: Trigger ArgoCD sync (optional)
-        if: inputs.environment == 'dev'
-        continue-on-error: true
+      - name: Deployment note
         run: |
-          echo "ArgoCD will auto-detect the change within 3 minutes"
-          echo "Or manually sync via UI: https://argocd-dev.openmeet.net"
+          echo "Committed new tag to k8s/environments/${{ inputs.environment }}."
+          echo "ArgoCD (do-prod app, automated sync) will roll this out within ~3 minutes."
+          echo "Watch: kubectl -n prod rollout status deployment/${{ inputs.service_name }}"