EDIT: The symptoms below are caused by an incomplete docker-compose.yml file. See #11763 (comment) for the solution.
I'm turning insane.
I have several functions deployed on a self hosted instance.
The function deployments are active, the build complete successfully.
[...]
Packing build ...
Build finished.
[14:48:37] [appwrite] Deployment finished.
Executions are created but never actually executed.
The function should be printing logs, should be finishing within seconds and returning data.
Status code 0, no logs, no errors, execution timing out after 30 minutes.
The runtime is booted up and ready.
The runtime is on the same network as the openruntimes-executor.
The containers on the appwrite network:
The openruntimes-executor logs show no errors.
The task-scheduler-functions container shows no activity.
Same with the task-scheduler-executions container.
The worker-functions container logs out "Fetched 0 functions" when an execution is created.
I'm on Appwrite 1.9.0, fresh volumes, have purged the configs several times, tried adding OPR_EXECUTOR_IMAGE out of pure desperation (still getting that same warning, don't think its even relevant as the env file doesn't contain it).
The docker compose file is taken directly from https://appwrite.io/install/compose, after which it was modified by the upgrade installer (the compose file present is still on 1.8.1, the installer from https://appwrite.io/docs/advanced/self-hosting/production/updates updated to 1.9.0 and updated the compose file).
At this point, there are zero error messages to be found anywhere, no hints as to what could be wrong, the runtime is being booted up, the deployment succeeds without any errors (in fact it shows success), the execution is created but code is never actually executed.
Judging by the fact that the function worker is printing "Fetched 0 functions" as soon as an execution is created shows that something is triggering, but it fetches 0 functions. How can I debug this further?
This is extremely frustrating.
EDIT: The symptoms below are caused by an incomplete docker-compose.yml file. See #11763 (comment) for the solution.
I'm turning insane.
I have several functions deployed on a self hosted instance.
The function deployments are active, the build complete successfully.
Executions are created but never actually executed.
The function should be printing logs, should be finishing within seconds and returning data.
Status code 0, no logs, no errors, execution timing out after 30 minutes.
The runtime is booted up and ready.
The runtime is on the same network as the openruntimes-executor.
The containers on the appwrite network:
The openruntimes-executor logs show no errors.
The task-scheduler-functions container shows no activity.
Same with the task-scheduler-executions container.
The worker-functions container logs out "Fetched 0 functions" when an execution is created.
I'm on Appwrite 1.9.0, fresh volumes, have purged the configs several times, tried adding OPR_EXECUTOR_IMAGE out of pure desperation (still getting that same warning, don't think its even relevant as the env file doesn't contain it).
The docker compose file is taken directly from https://appwrite.io/install/compose, after which it was modified by the upgrade installer (the compose file present is still on 1.8.1, the installer from https://appwrite.io/docs/advanced/self-hosting/production/updates updated to 1.9.0 and updated the compose file).
At this point, there are zero error messages to be found anywhere, no hints as to what could be wrong, the runtime is being booted up, the deployment succeeds without any errors (in fact it shows success), the execution is created but code is never actually executed.
Judging by the fact that the function worker is printing "Fetched 0 functions" as soon as an execution is created shows that something is triggering, but it fetches 0 functions. How can I debug this further?
This is extremely frustrating.