Mailcow-dockerized: SOGo creates random docker volume

[ntimo]

Hello,
I just noticed that SOGo after the update from today evening creates a random Docker volume this volume is not named with the mailcow name scheme. Even worse when mailcow gets restarted new volume is created and the old one is not deleted. Why does this happen, and why not use a static image name like mailcowdockerized_sogo-vol-1

[mkuron]

Caused by 8a88514. I guess it‘s correct to not use a named volume since it does not store persistent data. Also, using a named volume would cause trouble with SOGo updates as the contents of the volume might not get updated automatically.

However, it’s indeed not so nice that it doesn’t get cleaned up. Using tmpfs for this purpose would make sense, but tmpfs can’t be shared between containers. Maybe @andryyy has an idea how to solve this.

[andryyy]

I'll try to find a solution.

[ntimo]

Hello again,
@andryyy do you already have a solution for this :) My docker volume list is getting quite cluttered.

[stale[bot]]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

[ntimo]

Hello @andryyy are you still looking into this or can I/the bot close the issue? :)

[andryyy]

Still thinking about a solution. :)

[stale[bot]]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

[andryyy]

[ntimo]

@andryyy Could it be a solution to create a new volume called sogo-vol-1 and mount it to /usr/lib/GNUstep/SOGo/ inside of the SOGo and NGINX container? I tested this on my backup mx and it worked pretty well. I can create a PR if you want.

[andryyy]

It works as long as the web statics are not updated.

A very ugly fix would be to rsync the web docs to a sogo-web-vol-1 volume.

[ntimo]

Or maybe somehow (entrypoint?) delete the content of the volume every time the Sogo container is restarted.

[andryyy]

It is empty then, the content is not updated.

[ntimo]

As far as I know, when the SOGo container starts, it will generate all the files that are nessary. This would make it possible to regenerate them at container startup. We only need to remove them first.

[andryyy]

How would it do that? You need to reinstall SOGo to get those files back.

[andryyy]

As much as I hate myself for it: I created a named volume and rsync the 50 M of content...

We need to find a better fix in the future.