https://prefect.io logo
Join Slack
Powered by
<@ULVA73B9P> I'm having a trouble with compose.ym...
# marvin-ai
j
@Marvin I'm having a trouble with compose.yml aplication. All the services of my ccompose are up, but worker no! See the messagem on the ui: "Can't connect to Server API at http://0.0.0.0:4200/api. Check that it's accessible from your machine." This is my compose file: services: postgres: image: postgres:14 environment: POSTGRES_USER: prefect POSTGRES_PASSWORD: prefect POSTGRES_DB: prefect volumes: - postgres_data:/var/lib/postgresql/data healthcheck: test: ["CMD-SHELL", "pg_isready -U prefect"] interval: 5s timeout: 5s retries: 5 redis: image: redis:7 volumes: - redis_data:/data healthcheck: test: ["CMD-SHELL", "redis-cli ping"] interval: 5s timeout: 5s retries: 5 prefect-server: image: prefecthq/prefect:3-latest depends_on: postgres: condition: service_healthy redis: condition: service_healthy environment: PREFECT_API_DATABASE_CONNECTION_URL: postgresql+asyncpg://prefect:prefect@postgres:5432/prefect PREFECT_SERVER_API_HOST: 0.0.0.0 PREFECT_MESSAGING_BROKER: prefect_redis.messaging PREFECT_MESSAGING_CACHE: prefect_redis.messaging PREFECT_REDIS_MESSAGING_HOST: redis PREFECT_REDIS_MESSAGING_PORT: 6379 PREFECT_REDIS_MESSAGING_DB: 0 command: prefect server start --no-services ports: - "4200:4200" prefect-services: image: prefecthq/prefect:3-latest depends_on: postgres: condition: service_healthy redis: condition: service_healthy environment: PREFECT_API_DATABASE_CONNECTION_URL: postgresql+asyncpg://prefect:prefect@postgres:5432/prefect PREFECT_MESSAGING_BROKER: prefect_redis.messaging PREFECT_MESSAGING_CACHE: prefect_redis.messaging PREFECT_REDIS_MESSAGING_HOST: redis PREFECT_REDIS_MESSAGING_PORT: 6379 PREFECT_REDIS_MESSAGING_DB: 0 command: prefect server services start prefect-worker: image: prefecthq/prefect:3-latest depends_on: prefect-server: condition: service_started environment: PREFECT_API_URL: http://prefect-server:4200/api command: prefect worker start --pool local-pool volumes: postgres_data: redis_data:
m
✅ thought for 910.6 seconds
Thanks for sharing your compose file — a couple of small config tweaks should fix both the UI error and the worker issue. What’s happening - The UI message “Can’t connect to Server API at http://0.0.0.0:4200/api” means the browser is trying to talk to 0.0.0.0, which is only a bind address, not a routable address from your machine. - Your worker likely isn’t starting because it can’t reach the API or because the work pool 
local-pool
doesn’t exist yet. - I also see angle brackets 
<...>
around URLs in your env vars — if those are literally present, they’ll break connections. What to change 1) Remove angle brackets from URLs - Use plain strings like 
postgresql+asyncpg://...
and 
http://...
, not 
<postgresql+asyncpg://...>
or 
<http://...>
. 2) Tell the UI the correct external API URL - Keep the server binding on 
0.0.0.0
for Docker, but set the API URL that the UI should use from your browser. - Add 
PREFECT_API_URL: <http://localhost:4200/api>
to the 
prefect-server
service. This will make the UI talk to 
<http://localhost:4200/api>
instead of 
0.0.0.0
. 3) Make the worker talk to the server over the compose network - Keep 
PREFECT_API_URL: <http://prefect-server:4200/api>
in the worker (no angle brackets). 4) Ensure the worker starts after the API is ready and that a work pool exists - Add a healthcheck to 
prefect-server
and make the worker depend on it being healthy. - Create the work pool 
local-pool
with type 
process
once the API is up. Here’s a tightened compose snippet you can try
Copy code
services:
  postgres:
    image: postgres:14
    environment:
      POSTGRES_USER: prefect
      POSTGRES_PASSWORD: prefect
      POSTGRES_DB: prefect
    volumes:
      - postgres_data:/var/lib/postgresql/data
    healthcheck:
      test: ["CMD-SHELL", "pg_isready -U prefect"]
      interval: 5s
      timeout: 5s
      retries: 5

  redis:
    image: redis:7
    volumes:
      - redis_data:/data
    healthcheck:
      test: ["CMD-SHELL", "redis-cli ping"]
      interval: 5s
      timeout: 5s
      retries: 5

  prefect-server:
    image: prefecthq/prefect:3-latest
    depends_on:
      postgres:
        condition: service_healthy
      redis:
        condition: service_healthy
    environment:
      PREFECT_API_DATABASE_CONNECTION_URL: <postgresql+asyncpg://prefect:prefect@postgres:5432/prefect>
      PREFECT_SERVER_API_HOST: 0.0.0.0
      PREFECT_API_URL: <http://localhost:4200/api>
      PREFECT_MESSAGING_BROKER: prefect_redis.messaging
      PREFECT_MESSAGING_CACHE: prefect_redis.messaging
      PREFECT_REDIS_MESSAGING_HOST: redis
      PREFECT_REDIS_MESSAGING_PORT: 6379
      PREFECT_REDIS_MESSAGING_DB: 0
    command: prefect server start --no-services
    ports:
      - "4200:4200"
    healthcheck:
      test: ["CMD", "curl", "-sf", "<http://localhost:4200/api/health>"]
      interval: 5s
      timeout: 5s
      retries: 10

  prefect-services:
    image: prefecthq/prefect:3-latest
    depends_on:
      postgres:
        condition: service_healthy
      redis:
        condition: service_healthy
    environment:
      PREFECT_API_DATABASE_CONNECTION_URL: <postgresql+asyncpg://prefect:prefect@postgres:5432/prefect>
      PREFECT_MESSAGING_BROKER: prefect_redis.messaging
      PREFECT_MESSAGING_CACHE: prefect_redis.messaging
      PREFECT_REDIS_MESSAGING_HOST: redis
      PREFECT_REDIS_MESSAGING_PORT: 6379
      PREFECT_REDIS_MESSAGING_DB: 0
    command: prefect server services start

  prefect-worker:
    image: prefecthq/prefect:3-latest
    depends_on:
      prefect-server:
        condition: service_healthy
    environment:
      PREFECT_API_URL: <http://prefect-server:4200/api>
    command: prefect worker start --pool local-pool
    restart: unless-stopped

volumes:
  postgres_data:
  redis_data:
Create the work pool (run once) - After the API is healthy, create the pool the worker will poll:
Copy code
docker compose exec prefect-server prefect work-pool create local-pool --type process
Quick checks
- From your host: 
curl -sf <http://localhost:4200/api/health>
should return healthy JSON. - From the worker container: 
docker compose exec prefect-worker curl -sf <http://prefect-server:4200/api/health>
- Inside any container: 
prefect config view
should show 
PREFECT_API_URL
as you set it. References - Docker compose guide and self-hosted server notes: Prefect 3: Self-hosted with Docker Compose and Prefect 3: Server in Docker If you try the above and still see the UI error, let me know: - the exact URL you open in the browser - the output of 
curl -sf <http://localhost:4200/api/health>
- the first 50 lines of logs from 
prefect-server
and 
prefect-worker
(
docker compose logs --no-color --tail=50 <service>
)
j
@Marvin The container prefect-container-deploy-prefect-worker-1 had error see the log*:*
Copy code
File "/usr/local/lib/python3.12/site-packages/prefect/client/orchestration/base.py", line 53, in request
    return await self._client.send(request)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.12/site-packages/prefect/client/base.py", line 348, in send
    response = await self._send_with_retry(
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.12/site-packages/prefect/client/base.py", line 268, in _send_with_retry
    response = await send(request, *send_args, **send_kwargs)
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.12/site-packages/httpx/_client.py", line 1629, in send
    response = await self._send_handling_auth(
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.12/site-packages/httpx/_client.py", line 1657, in _send_handling_auth
    response = await self._send_handling_redirects(
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.12/site-packages/httpx/_client.py", line 1694, in _send_handling_redirects
    response = await self._send_single_request(request)
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.12/site-packages/httpx/_client.py", line 1730, in _send_single_request
    response = await transport.handle_async_request(request)
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.12/site-packages/httpx/_transports/default.py", line 393, in handle_async_request
    with map_httpcore_exceptions():
         ^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.12/contextlib.py", line 158, in __exit__
    self.gen.throw(value)
  File "/usr/local/lib/python3.12/site-packages/httpx/_transports/default.py", line 118, in map_httpcore_exceptions
    raise mapped_exc(message) from exc
httpx.ConnectError: All connection attempts failed
I solved running prefect server update, see: docker compose exec prefect-server prefect server database upgrade After the error on interface keep with message: Can't connect to Server API at http://0.0.0.0:4200/api. Check that it's accessible from your machine. I entry inside prefect-container-deploy-prefect-worker-1 and executed the command prefect config view: And I have: 🚀 you are connected to: http://prefect-server:4200 PREFECT_PROFILE='ephemeral' PREFECT_API_URL='http://prefect-server:4200/api' (from env) PREFECT_SERVER_ALLOW_EPHEMERAL_MODE='true' (from profile) I made a curl command: curl http://prefect-server:4200/api/health inside prefect-container-deploy-prefect-worker-1 container and the response was "true". After made docker compose up -d, all the containers be running, but the message error keep on interface
m
✅ thought for 172.3 seconds
Thanks for all the details — they’re super helpful. What’s happening - Your worker can reach the API just fine on the Docker network (`http://prefect-server:4200/api`), confirmed by 
curl .../health
. - The browser UI, though, is trying to talk to 
<http://0.0.0.0:4200/api>
, which is a bind address and not reachable from your machine. That value comes from the server’s UI settings. Fix You need to tell the Prefect UI which externally reachable API URL to use by setting 
PREFECT_UI_API_URL
on the server container to the exact URL you use in your browser (usually 
<http://localhost:4200/api>
on your dev machine, or your host/IP/domain in other setups). Example docker-compose snippet Update your prefect-server service to include these env vars (adjust host as needed):
Copy code
services:
  prefect-server:
    image: prefecthq/prefect:3-latest
    command: prefect server start
    ports:
      - "4200:4200"
    environment:
      # What the UI should use to reach the API from your browser
      PREFECT_UI_API_URL: "<http://localhost:4200/api>"
      # Optional but recommended for correct link generation in emails/UI
      PREFECT_UI_URL: "<http://localhost:4200>"
Then restart:
Copy code
docker compose up -d --force-recreate
Verify - From your browser: open http://localhost:4200/ui-settings and check that 
api_url
matches 
<http://localhost:4200/api>
. - From your machine: 
curl <http://localhost:4200/api/health>
should return 
true
. - Hard refresh the UI (clear cache or Ctrl/Cmd+Shift+R). Notes - Inside Docker, your worker should keep using 
PREFECT_API_URL=<http://prefect-server:4200/api>
(the service name). The browser UI must use a host that your machine can reach (e.g., 
localhost
, your host IP, or a domain). - If you’re accessing the UI from another machine or behind a reverse proxy, set 
PREFECT_UI_API_URL
to that external address (e.g., 
<http://your-host:4200/api>
or your proxy path like `https://example.com/prefect/api`). If you can share your docker-compose.yml (the 
prefect-server
service, especially ports and environment), I can tailor the exact values. References - UI and security settings: Prefect 3 Security &amp; UI settings - Settings models reference (UI/API): Prefect 3 Settings models
3 Views
Open in Slack
PreviousNext
Powered by