I’ve used Nextcloud for quite a while and updated it more times than I can count, but never had this issue.
Got a notification yesterday there’s an update, but when I click the updater, it goes to a NC page that says there’s a 404. It gives me the option to download a zip file, but I’ve never updated this way… I can reach Nextcloud via my domain name which is reverse proxy’d through swag… so I know NC is online. Same thing happens if I try to upgrade locally via NC’s ip:port. I repulled NC this morning to make sure I had the latest image, and I do.
Thanks for the answer… but this seems like a backwards change that just complicates matter.
Anyway, so I changed the image tag on my nextcloud compose to the new version (27.0.2) now, for whatever reason, I get a mariadb error when I try to redeploy.
failed to pull images of the stack: nextcloud Pulling mariadb Pulling mariadb Pulled nextcloud Error Error response from daemon: manifest unknown
paste code here
the change wasn’t for people who can read, it was made for unraid users. Take a look at our discord threads at how many unraid users have borked nextcloud setups. this is to ease OUR burden, not yours. Though it will immensely help many unraid users.
Seems like you were probably one of the few users that would keep their version of Nextcloud maintained correctly
We decided to make the change to streamline the updates (as mentioned by driz) 80% of our support requests were how to update a container and users not following updates (some were even still on v19 at the point of v27 being released)
Fair enough, but it sounds like you’ve inconvenienced/made things more difficult on a bunch of people, to accommodate a few on Unraid who didn’t understand how to keep their systems update.
unfortunately, it is far more than a few unraid users… but you are right, it will inconvenience some users, especially those who were using release candidates rather than official releases.
It might be something we can look to address, but I would not put an ETA on that
it’s worth also noting that nextcloud (and plex) are our few containers with a different upgrade process, this change also brings NC in line with how 99% of our containers work, but again, i get the displeasure with the change, it negatively affects me too, though I’m not a big users of NC
Yeah… If I’m sounding critical, it’s not really intended. I’ve always had pretty good luck with LS containers and NC has been no exception. As said above… I keep anything I have facing the Internet regularly updated, NC no exception,.
I never had issues before, so I guess that’s why this change kinda irks me (don’t use Plex, so can’t really speak on it)
I’ll be honest… I just done my first update like this. I’m not happy at all. I have never had issues upgrading NC… this time… I redeployed my compose… NC is failing to start… Log isn’t showing anything.
I check my config.php, it’s now completely blank…
I’ll check this in the morning but it looks like this has caused me quite a freaking mess
Yeah, finally got the container reinstalled, and it’s telling me NC is not installed… Bashed into the container, none of my settings are there, config.php is empty…
Well, thankfully, I keep thorough backups and actually had jus backed up all my stuff Sunday afternoon, so I lost nothing.
First, I tried to just delete my containers and redeploy my compose. That didn’t work as it reported my container needed to be configured… so somehow this nuked my data folders as well (config for nextcloud and mariadb, and then my nextcloud data folder)
So I deleted the containers.
Restored those 3 folders from my backup back to their proper position
Redeployed NC, and it appears all is right in the world again.