jenkins server runs out of disk space

[Nov11]

Build processes fail because of this. Need help to make more valid disk space on that server.

[crd477]

Thanks. I'll fix this.

[crd477]

I've implemented a workaround in a cron job on each of the workers that removes workspaces from old (>1 day) builds. The correct long term fix is to have the proper cleanup stage in the Jenkins pipeline. I'll test the solution and submit a PR when that is ready.

[tcm-marcel]

@crd477 Does the workspace also include the logs? Because in general we want to keep them as long as possible, at least one week. We had the problem already that we wanted to analyze older builds and the logs were gone. If you then click on the Jenkins link of a commit on Github, you just get a 404. The cleanup stage you mentioned seems what we want: delete everything except for the logs and then keep the workspace for a while.

[crd477]

On 2018-04-17 14:12, Marcel Kost wrote:

@crd477 https://github.com/crd477 Does the workspace also include the logs? Because in general we want to keep them as long as possible, at least one week. We had the problem already that we wanted to analyze older builds and the logs were gone. If you then click on the Jenkins link of a commit on Github, you just get a 404. The cleanup stage you mentioned seems what we want: delete everything except for the logs and then keep the workspace for a while.

I'm pretty sure the cron job only removes the actual source and object code from a particular build, not the logs and results. I wonder if the 404 you saw was from an aborted build or something. Could you point me to an example so I can check it out?

Thanks...

[crd477]

Also of note, the peloton Jenkins project is configured to keep old items for 30 days.

[tcm-marcel]

@crd477 It was some weeks ago, I can't find it anymore. Probably it got caused by Jenkins changes at that time. I will tell you, if it happens again.