Full disk issues

[anbraten]

We had a full disk as the woodpecker server was set to trace log-level and filled the disk.

Several problems brought us to this incident:

• ansible was not updating the container as the log-level was changed (probably sth related to has to be changed)
• we had no limit for the log files: https://docs.docker.com/engine/logging/drivers/json-file/ => a global default limit should be set
• old container images (especially the renovate updates of the woodpecker-server / autoscaler) were never deleted => Add docker prune to drop unused resources regularly #246

[xoxys]

Just switch to journald logger and let the pre-configured journal looging handle it.

[6543]

I would say just needs a pull ☝️ now?