Please, before posting, we’ve seen a lot of these post saying X instance is down right now, or “is X instance down?”. The answer is probably, but it’s probably because Lemmy v0.19 dropped and it includes some downtime, upwards of an hour for all the migrations to finish.
Please, hold off on asking/posting for an hour or so. Your sysadmins are waiting for it to come back online.
I manage my own tiny insurance and it took long enough that I thought I broke something. I must need to turn up logging because the back end was completely silent except for “0.19.0” and I only knew something was happening because the database was busy.
Same happened to me, I learned on other releases that no news is good news usually in updates, if you see things busy it’s probably doing stuff - but this was a nail biter for sure. Very glad I skipped all the RCs.
I chose to take down my production instance of lemmyz do a cp -r of it, update the docker compose file to specify the new version numbers for Lemmy and lemmy-ui plus change the port that the stack exposes and do a docker-compose pull && docker-compose up -d on that stack to check to see if it could all work. Thankfully it did, so I updated the production compose file and tada! I’m on 19.0 safely
Did it in a similar way. I used:
to keep permissions also. Worked like a charm.
Ah, there’s a flag I’ve never used before. That’s been noted and I’ll do the same thing myself in the future to minimise risk of breakage in the future.
Are you using disk storage or object storage for pict-rs? I need a better way to test updates but I’m trying to avoid having to duplicate my storage or risk breaking prod by pointing to my live object storage from a second instance.
I’m just using some of the spare HDD space to store it all. If there was more than one user on my instance I’d probably need to find a better method, but for me, for now, it works well enough.