search

dirtsimple / poste.io

poste.io mail server, but with IP management and custom webmail plugins
153 stars 28 forks source link

how to clear logs #28

Open flourishing opened 1 year ago

flourishing commented 1 year ago

I tried to make migration of poste.io. when i tar it, too many logs. it tooks about one hour to pack it. how to clear the logs. and which logs or directory or files we can delete them ? thanks .

pjeby commented 1 year ago

I don't understand what you mean by "make migration". All the data and configuration is stored in the data volume, so that should be all you need. Logs are in the log subdirectory of the data volume, and are automatically rotated, unless you're using a really old version of poste. (There is an old version that would accumulate tons and tons of logs on a daily basis due to failed rotations.)

flourishing commented 1 year ago

sorry . i mean I try to move poste docker container to another server. when i try to backup the data volume, it has many many log files. it take me one hour. i used the latest docker image, which downloaded about 7 months ago. so i don't know is it too old or not . I will try to upgrade to the newest version. see what happend. thanks for the response quickly.

flourishing commented 1 year ago

hi, i have tried to install a new one. i met a problem, got an error message. and can't start the container. 023-05-24 20:09:33 #676(main) <79a2dc>; main; main: rspamd 3.1 is loading configuration, build id: release 2023-05-24 20:09:33 s6-svscan: fatal: directory already locked 2023-05-24 20:09:33 [cmd] s6-svscan exited 100 2023-05-24 20:09:33 [cont-finish.d] executing container finish scripts... 2023-05-24 20:09:33 [cont-finish.d] done. 2023-05-24 20:09:33 [s6-finish] syncing disks. 2023-05-24 20:09:34 [s6-finish] sending all processes the TERM signal. 2023-05-24 20:09:37 [s6-finish] sending all processes the KILL signal and exiting.

pjeby commented 1 year ago

So, this was the log rotation issue: https://bitbucket.org/analogic/mailserver/issues/818/roundcube-logs-are-rotated-recursively - it was fixed as of 2.2.27, but if you ever used an older version you would likely have excess logs even after the problem was fixed. You'll want to check and see if there's a huge roundcube/logs directory under the data volume, and delete those files.

With regard to the "directory already locked" problem, my guess would be that you backed up the data volume while the mail server was running and thus have a lock file present? Probably you will need to do the backup with the mail server offline. I was not able to find anything on Google about that error, so I don't know for sure.