Use seperate container for running playwright

Author: thenav56

Dockerfile

@@ -33,9 +33,6 @@ RUN --mount=type=cache,target=$UV_CACHE_DIR \
     --mount=type=bind,source=pyproject.toml,target=pyproject.toml \
     uv sync --frozen --no-install-project --all-groups
 
-RUN playwright install \
-    && playwright install-deps
-
 # To avoid some SyntaxWarnings ("is" with a literal), still needed on 20241024:
 ENV AZUREROOT=/usr/local/lib/python3.11/site-packages/azure/storage/
 RUN perl -pi -e 's/ is 0 / == 0 /'      ${AZUREROOT}blob/_upload_chunking.py

api/tasks.py

@@ -55,6 +55,7 @@ def generate_url(url, export_id, user, title):
     export = Export.objects.get(id=export_id)
     user = User.objects.get(id=user)
     token = Token.objects.filter(user=user).last()
+    logger.info(f"Starting export: {export.pk}")
 
     footer_template = """
         <div class="footer" style="width: 100%;font-size: 8px;color: #FEFEFE; bottom: 10px; position: absolute;">
@@ -86,19 +87,7 @@ def generate_url(url, export_id, user, title):
     try:
         with tempfile.TemporaryDirectory() as tmp_dir:
             with sync_playwright() as p:
-                browser = p.chromium.launch(
-                    headless=True,
-                    args=[
-                        # https://github.com/microsoft/playwright-python/issues/1453
-                        # Usually required when running AWS lambda
-                        # "--single-process",
-                        "--no-zygote",
-                        "--no-sandbox",
-                        "--disable-setuid-sandbox",
-                        "--disable-dev-shm-usage",
-                    ],
-                    devtools=False,
-                )
+                browser = p.chromium.connect(settings.PLAYWRIGHT_SERVER_URL)
                 storage_state = build_storage_state(
                     tmp_dir,
                     user,
@@ -108,7 +97,8 @@ def generate_url(url, export_id, user, title):
                 page = context.new_page()
                 if settings.DEBUG_PLAYWRIGHT:
                     DebugPlaywright.debug(page)
-                timeout = 300000
+                # FIXME: Use of Timeout correct?
+                timeout = 300_000  # 5 min
                 page.goto(url, timeout=timeout)
                 time.sleep(5)
                 page.wait_for_selector("#pdf-preview-ready", state="attached", timeout=timeout)
@@ -143,3 +133,4 @@ def generate_url(url, export_id, user, title):
         )
         export.status = Export.ExportStatus.ERRORED
         export.save(update_fields=["status"])
+    logger.info(f"End export: {export.pk}")

docker-compose.yml

@@ -19,6 +19,7 @@ x-server: &base_server_setup
     GO_ENVIRONMENT: ${GO_ENVIRONMENT:-development}
     API_FQDN: ${API_FQDN:-http://localhost:8000}
     FRONTEND_URL: ${FRONTEND_URL:-http://localhost:3000}
+    PLAYWRIGHT_SERVER_URL: ${PLAYWRIGHT_SERVER_URL:-ws://playwright:3000/}
     GO_WEB_INTERNAL_URL: ${GO_WEB_INTERNAL_URL:-http://host.docker.internal:3000}
     DJANGO_ADDITIONAL_ALLOWED_HOSTS: ${DJANGO_ADDITIONAL_ALLOWED_HOSTS:-host.docker.internal}
     DEBUG_EMAIL: ${DEBUG_EMAIL:-true}
@@ -57,7 +58,7 @@ x-server: &base_server_setup
     - db
     - redis
     - elasticsearch
-
+    - playwright
 
 services:
   db:
@@ -76,7 +77,15 @@ services:
     volumes:
       - redis-data:/data
 
-  # NOTE: Used Only for local development
+  # NOTE: Make sure this matches with pyproject playwright dependency
+  playwright:
+    image: mcr.microsoft.com/playwright:v1.50.0-noble
+    working_dir: /home/pwuser
+    user: pwuser
+    command: bash -c "npx -y [email protected] run-server --port 3000 --host 0.0.0.0"
+    extra_hosts:
+      - "host.docker.internal:host-gateway"
+
   elasticsearch:
     image: docker.elastic.co/elasticsearch/elasticsearch:7.0.0
     container_name: elasticsearch
@@ -96,7 +105,6 @@ services:
     ports:
       - 9200:9200
 
-  # NOTE: Used Only for local development
   kibana:
     image: 'docker.elastic.co/kibana/kibana:7.0.0'
     container_name: kibana

main/runserver.sh

@@ -9,6 +9,7 @@ python manage.py collectstatic --noinput -l
 #python manage.py make_permissions
 
 # Add server name(s) to django settings and nginx - later maybe only nginx would be enough, and ALLOWED_HOSTS could be "*"
+# TODO: Allow API_FQDN with https
 if [ "$GO_ENVIRONMENT"x = productionx ]; then
     sed -i 's/\$NGINX_SERVER_NAME/'$API_FQDN' api.go.ifrc.org/g' /etc/nginx/sites-available/nginx.conf
     sed -i 's/CHANGE_ME_BEFORE_START/prod/' /etc/nginx/sites-available/nginx.conf

main/settings.py

@@ -33,6 +33,7 @@
     API_FQDN=str,  # https://goadmin.ifrc.org
     FRONTEND_URL=str,  # https://go.ifrc.org
     GO_WEB_INTERNAL_URL=(str, None),  # http://host.docker.internal
+    PLAYWRIGHT_SERVER_URL=str,  # ws://playwright:3000/
     # Database
     DJANGO_DB_NAME=str,
     DJANGO_DB_USER=str,
@@ -173,6 +174,8 @@ def parse_domain(*env_keys: str) -> str:
 GO_WEB_INTERNAL_URL = parse_domain("GO_WEB_INTERNAL_URL", "FRONTEND_URL")
 FRONTEND_URL = urlparse(GO_WEB_URL).hostname  # FIXME: Deprecated. Slowly remove this from codebase
 
+PLAYWRIGHT_SERVER_URL = env("PLAYWRIGHT_SERVER_URL")
+
 INTERNAL_IPS = ["127.0.0.1"]
 if env("DOCKER_HOST_IP"):
     INTERNAL_IPS.append(env("DOCKER_HOST_IP"))