deB4SH

Does someone has an idea why gitea action recently got issues with complete job phase? The job itself runs fine. On the action host all containers are stopped and properly decommissioned but the job seems to be stuck and ends with an error after some time.

tags for visibility: #gitea #GiteaActions #selfhosted

Sep 15, 2025 at 5:30pmWeb
Show threadLouisSep 15, 2025
@deb4sh what kind of job fails? All of them or only certain kind of jobs/operations?
Show threaddeB4SHSep 15, 2025
@louis all of them fail in the complete phase. Feels like the act runner loses the track of the docker containers created for the runs. Deletes them accordingly but misses out on them being removed.
Show threadLouisSep 15, 2025
@deb4sh just to be sure, both Gitea server and action are up to date? Nothing in the server logs and nothing in the action runner logs, nor in the kernel/console logs for the machine hosting the runner?
Could also feel like a memory issue maybe?
Show threaddeB4SHSep 18, 2025

@louis tested a bit further, feels like this issue occures when a docker container is build. Seems like the service is stuck on cleaning up the environment afterwards.

```
Cleaning up container for job build-doku
Removed container: 766e6f90fb6ea617a6eca69f26607dd94344f9d7ad70f3d48a7de8c6fcf5a41c
🐳 docker volume rm GITEA-ACTIONS-TASK-3755_WORKFLOW-build_JOB-build-doku
```

The volume is gone. Checked via cli. Hmm. Weird issue