update the backend_build_script

.gitea/workflows/backend-ecr-modal.yaml

@@ -1,52 +1,29 @@
-# Gitea Actions Workflow: Backend Build via Modal (Serverless)
+# Gitea Actions Workflow: Backend ECR Deployment
-# Place at: .gitea/workflows/backend-modal.yaml
+# Place at: .gitea/workflows/backend-ecr.yaml (in repository root)
-#
-# This workflow offloads the heavy Docker build to Modal's serverless infrastructure,
-# avoiding resource constraints on the 2GB RAM Lightsail VM runner.
-#
-# Required Gitea Repository Secrets:
-#   MODAL_TOKEN          - Modal API token (from modal.com/settings)
-#
-# Required Modal Secrets (configured in Modal dashboard, NOT Gitea):
-#   aws-secrets          - AWS_ACCESS_KEY_ID, AWS_SECRET_ACCESS_KEY, AWS_REGION
-#
-# Required Gitea Repository Variables:
-#   ECR_PUBLIC_REGISTRY_ALIAS - Your public.ecr.aws alias (e.g., vkist-project)
-#   TRITON_ENDPOINT          - Modal Triton endpoint for CV inference
-#   CV_INFERENCE_HOST        - Bind host (default: 0.0.0.0)
-#   CV_INFERENCE_PORT        - Bind port (default: 8001)
-#   BASE_URL                 - External base URL for routing
 
-name: Backend Modal Build
+name: Backend EC - Modal  Deployment
 
 on:
+  push:
+    branches: ['**']
   workflow_dispatch: {}
-  # push:
-  #   branches: ['**']
-  # workflow_dispatch:
-  #   inputs:
-  #     tag:
-  #       description: 'Image tag (default: git SHA)'
-  #       required: false
-  #       type: string
-  #     platform:
-  #       description: 'Target platform'
-  #       required: false
-  #       type: string
-  #       default: 'linux/amd64'
 
 permissions:
   contents: read
   packages: write
 
 env:
+  # Build-time / workflow context
   BACKEND_DIR: workspace/sprint_1_2/CODEBASE/backend
   ECR_REGISTRY: public.ecr.aws
   ECR_REPOSITORY: ${{ vars.ECR_PUBLIC_REGISTRY_ALIAS }}/msk-cv-inference-server
   DOCKERFILE_PATH: deps/implementation/backend_deploy/Dockerfile
   MODAL_SCRIPT: workspace/sprint_1_2/CODEBASE/deps/implementation/backend_deploy/gitea_modal_build.py
 
-  # Runtime config (forwarded to container via deployment)
+  # CV inference server runtime configuration
+  # These are Gitea repository variables (or secrets) that are forwarded
+  # to the container at deploy time. The Docker image itself does NOT bake
+  # them in — the FastAPI app reads them from the environment via pydantic-settings.
   TRITON_ENDPOINT: ${{ vars.TRITON_ENDPOINT }}
   CV_INFERENCE_HOST: ${{ vars.CV_INFERENCE_HOST }}
   CV_INFERENCE_PORT: ${{ vars.CV_INFERENCE_PORT }}
@@ -54,18 +31,42 @@ env:
   CORS_ORIGINS: ${{ vars.CORS_ORIGINS }}
   ECR_PUBLIC_REGISTRY_ALIAS: ${{ vars.ECR_PUBLIC_REGISTRY_ALIAS }}
 
+
 jobs:
-  build-via-modal:
+  build-and-push:
-    name: Build & Push via Modal
+    runs-on: ubuntu-latest
-    runs-on: [self-hosted, linux, amd64]
     container:
       image: gitea-modal-runner:latest 
-    timeout-minutes: 20
+    timeout-minutes: 30
-    
     steps:
       - name: Checkout repository
         uses: actions/checkout@v4
 
+      - name: Lint backend code
+        run: |
+          python -m py_compile workspace/sprint_1_2/CODEBASE/backend/cv_inference_server.py
+          find workspace/sprint_1_2/CODEBASE/backend -name '*.py' -exec python -m py_compile {} +
+
+      - name: Extract metadata for Docker
+        id: meta
+        uses: docker/metadata-action@v5
+        with:
+          # Full ECR Public image reference: public.ecr.aws/{alias}/{repo}
+          images: ${{ env.ECR_REGISTRY }}/${{ env.ECR_REPOSITORY }}
+          tags: |
+            type=ref,event=branch
+            type=sha,prefix=
+            type=raw,value=latest,enable={{is_default_branch}}
+          labels: |
+            org.opencontainers.image.source=${{ github.server_url }}/${{ github.repository }}
+            org.opencontainers.image.revision=${{ github.sha }}
+            org.opencontainers.image.title=VKIST CV Inference Server
+
+      - name: Print image URI and digest
+        run: |
+          echo "Pushed ${{ env.ECR_REGISTRY }}/${{ env.ECR_REPOSITORY }}:${{ steps.meta.outputs.version }}"
+          echo "Digest: ${{ steps.meta.outputs.tags }}"
+
       - name: Determine image tag
         id: tag
         run: |
@@ -90,15 +91,8 @@ jobs:
             --platform linux/amd64 \
             --push
 
-  # Optional: Trigger dependent deployments
-  # trigger-triton:
-  #   needs: build-via-modal
-  #   if: success() && github.event_name == 'push'
-  #   uses: ./.gitea/workflows/trigger_modal_triton.yaml
-  #   secrets: inherit
-
   notify-deploy:
-    needs: build-via-modal
+    needs: modal-build
     if: success() && github.event_name == 'push'
     runs-on: ubuntu-latest
     steps:

.gitea/workflows/backend-ecr.yaml

@@ -1,133 +0,0 @@
-# Gitea Actions Workflow: Backend ECR Deployment
-# Place at: .gitea/workflows/backend-ecr.yaml (in repository root)
-
-name: Backend ECR Deployment
-
-on:
-  push:
-    branches: ['**']
-  workflow_dispatch: {}
-
-permissions:
-  contents: read
-  packages: write
-
-env:
-  # Build-time / workflow context
-  BACKEND_DIR: workspace/sprint_1_2/CODEBASE/backend
-  ECR_REGISTRY: public.ecr.aws
-  ECR_REPOSITORY: ${{ vars.ECR_PUBLIC_REGISTRY_ALIAS }}/msk-cv-inference-server
-  DOCKERFILE_PATH: deps/implementation/backend_deploy/Dockerfile
-
-  # CV inference server runtime configuration
-  # These are Gitea repository variables (or secrets) that are forwarded
-  # to the container at deploy time. The Docker image itself does NOT bake
-  # them in — the FastAPI app reads them from the environment via pydantic-settings.
-  TRITON_ENDPOINT: ${{ vars.TRITON_ENDPOINT }}
-  CV_INFERENCE_HOST: ${{ vars.CV_INFERENCE_HOST }}
-  CV_INFERENCE_PORT: ${{ vars.CV_INFERENCE_PORT }}
-  BASE_URL: ${{ vars.BASE_URL }}
-  CORS_ORIGINS: ${{ vars.CORS_ORIGINS }}
-
-jobs:
-  build-and-push:
-    runs-on: ubuntu-latest
-    container:
-      image: gitea-modal-runner:latest 
-    timeout-minutes: 30
-    steps:
-      - name: Checkout repository
-        uses: actions/checkout@v4
-
-      - name: Lint backend code
-        run: |
-          python -m py_compile workspace/sprint_1_2/CODEBASE/backend/cv_inference_server.py
-          find workspace/sprint_1_2/CODEBASE/backend -name '*.py' -exec python -m py_compile {} +
-
-      # - name: Configure AWS credentials
-      #   uses: aws-actions/configure-aws-credentials@v4
-      #   with:
-      #     aws-access-key-id: ${{ secrets.AWS_ACCESS_KEY_ID }}
-      #     aws-secret-access-key: ${{ secrets.AWS_SECRET_ACCESS_KEY }}
-      #     aws-region: us-east-1
-
-      # - name: Login to Amazon ECR Public
-      #   id: login-ecr
-      #   uses: aws-actions/amazon-ecr-login@v2
-      #   with:
-      #     registry-type: public
-      #     mask-password: 'true'
-
-      - name: Extract metadata for Docker
-        id: meta
-        uses: docker/metadata-action@v5
-        with:
-          # Full ECR Public image reference: public.ecr.aws/{alias}/{repo}
-          images: ${{ env.ECR_REGISTRY }}/${{ env.ECR_REPOSITORY }}
-          tags: |
-            type=ref,event=branch
-            type=sha,prefix=
-            type=raw,value=latest,enable={{is_default_branch}}
-          labels: |
-            org.opencontainers.image.source=${{ github.server_url }}/${{ github.repository }}
-            org.opencontainers.image.revision=${{ github.sha }}
-            org.opencontainers.image.title=VKIST CV Inference Server
-
-      - name: Build and push Docker image
-        uses: docker/build-push-action@v5
-        with:
-          context: ./${{ env.BACKEND_DIR }}
-          file: ./${{ env.DOCKERFILE_PATH }}
-          push: true
-          platforms: linux/amd64,linux/arm64
-          tags: ${{ steps.meta.outputs.tags }}
-          labels: ${{ steps.meta.outputs.labels }}
-          cache-from: type=gha
-          cache-to: type=gha,mode=max
-
-      - name: Print image URI and digest
-        run: |
-          echo "Pushed ${{ env.ECR_REGISTRY }}/${{ env.ECR_REPOSITORY }}:${{ steps.meta.outputs.version }}"
-          echo "Digest: ${{ steps.meta.outputs.tags }}"
-
-  # trigger-triton:
-  #   needs: build-and-push
-  #   if: success() && github.event_name == 'push'
-  #   runs-on: ubuntu-latest
-  #   steps:
-  #     - name: Trigger Modal Triton deployment
-  #       uses: gitea/gitea-actions-workflow-trigger@v1
-  #       with:
-  #         repository: ${{ github.repository }}
-  #         workflow: trigger_modal_triton.yaml
-  #         ref: ${{ github.ref }}
-  #         inputs: '{}'
-
-  notify-deploy:
-      needs: build-and-push
-      if: success() && github.event_name == 'push'
-      runs-on: ubuntu-latest
-      steps:
-        - name: Trigger deployment webhook
-          if: env.DEPLOY_WEBHOOK_URL != ''
-          env:
-            DEPLOY_WEBHOOK_URL: ${{ secrets.DEPLOY_WEBHOOK_URL }}
-            WEBHOOK_SECRET: ${{ secrets.WEBHOOK_SECRET }}
-          run: |
-            cat <<EOF | curl -X POST "${DEPLOY_WEBHOOK_URL}" \
-              -H "Content-Type: application/json" \
-              -H "X-Gitea-Signature: sha256=$(echo -n '@-' | openssl dgst -sha256 -hmac "${WEBHOOK_SECRET}" | cut -d' ' -f2)" \
-              -d @-
-            {
-              "service": "cv-inference-server",
-              "image": "${{ env.ECR_REGISTRY }}/${{ env.ECR_REPOSITORY }}:${{ steps.meta.outputs.version }}",
-              "env": {
-                "TRITON_ENDPOINT": "${{ vars.TRITON_ENDPOINT }}",
-                "CV_INFERENCE_HOST": "${{ vars.CV_INFERENCE_HOST }}",
-                "CV_INFERENCE_PORT": "${{ vars.CV_INFERENCE_PORT }}",
-                "BASE_URL": "${{ vars.BASE_URL }}",
-                "CORS_ORIGINS": "${{ vars.CORS_ORIGINS }}",
-                "ECR_PUBLIC_REGISTRY_ALIAS": "${{ vars.ECR_PUBLIC_REGISTRY_ALIAS }}"
-              }
-            }
-            EOF