runtime.gunicorn_worker_timeout

gunicorn kills workers by timeout

A mevcore app hosting error at the “start / health check” stage.

What it means

A gunicorn worker did not answer within --timeout (30 s by default): a slow start (models loading) or a long request.

The build passed, but the app did not start or did not answer the health check.

Typical causes

The app does not listen on 0.0.0.0 and the port from the PORT variable
An environment variable is missing, or the database or Redis is not connected
The start command is wrong or the process exits right away

How to fix it on mevcore

Raise --timeout (for example 120) in the start command in the settings, use fewer workers, or add --preload.

1 Open the app in the mevcore panel: the Problems card on the overview shows this error and the log line it was found in.
It is fixed on the “Settings” tab of the app.
2 After the fix, deploy again. If it fails again, the log ends with [mevcore] lines naming the new error code.
3 The "Ask AI" button next to the error prepares a request for ChatGPT or Claude: the error code, the end of the log without secrets and the app settings.

For an AI assistant

The same in English: paste it together with the log into the AI chat.

mevcore error code: runtime.gunicorn_worker_timeout
phase: start / health check
diagnosis: gunicorn kills workers by timeout
platform hint: A gunicorn worker did not answer within --timeout (30 s by default): a slow start (models loading) or a long request. Raise --timeout (for example 120) in the start command in the settings, use fewer workers, or add --preload.
platform docs: https://mevcore.com/llms-full.txt

Can't fix it?

Message support with the error code: we will tell you what to change in the project or settings.

App hosting