>self updating function of nextcloud works very well

Then that's a new development. I've been using it since about v9 and it was a complete trainwreck that might have had a 25% success rate until I gave it up and moved to the docker around v17.

> Then that's a new development

my non-docker install has been going strong for 3-4 years now, so that's hardly what I call recent...

Yeah I run it baremetal. No issues at all.

I’ve had it bork my install so many times that in the end I spent more time recovering nextcloud than actually using it.

My instance is dead now, after I failed to recover it the last time and couldn’t be bothered anymore.

YMMV, but I’m out.

> I’ve had it bork my install so many times that in the end I spent more time recovering nextcloud than actually using it.

I had issues with upgrading a few times on a modest VPS, when trying to upgrade via the web interface. I've since switched to upgrading via ssh by running the `updater.phar` script[0] and haven't experienced an issue upgrading since.

I of course don't know if this would've avoided the issues you experienced, @josteink, but I wanted to mention it in case others have a similar problem to what I had.

[0] https://docs.nextcloud.com/server/latest/admin_manual/mainte...

That sucks. I'd be frustrated too. For me it just hosts calendar and contacts for a few users and my files are in seafile so perhaps the small footprint helps. In what way did it break, I want to keep an eye out for this.

It was a custom-built LXC container I built on an Alpine Linux base.

I used the Alpine packages to upgrade it, then afterwards I used the Nextcloud admin scripts to migrate the schema, apps & plugins.

Biggest clusterfuck I’ve ever dealt with. Not doing that again.

TBF the Alpine-setup probably made everything worse, and that’s a lesson learned, but I’m just fed up and can’t bother setting up a new instance now.