lemmy.cafe seems down. it returns a 502, indicating that something behind nginx went wrong. i checked !general@lemmy.cafe, but the last post is ~2 years ago. does anybody know whether .cafe has some planned downtime, or something else announced?
just prodding around, i also typed https://status.lemmy.cafe/ into my browser. that leads to their grafana login :-)
thanks to everyone contributing. but especially @Illecors@lemmy.cafe for not only maintaining the instance but also swiftly bringing it back up!
Things should be back.
There’s a problem I never got to figure out where the internal nginx container - sitting between anubis and UI containers - does not come back up after a restart.
The VM this was running on had a fault on DigitalOcean side and got moved to another host. This included booting it back up, which did not bring that nginx container up. API worked, so all apps worked. It’s just the standard, photon and voyager UIs were down.
Apologies.
Does the container not start at all or does it crash? I forgot the specifics but nginx is kinda fragile when the site it’s supposed to proxy to is not ready
It starts just fine, just doesn’t come up on its own after boot. If you’re interested - have a look https://gitlab.com/MithrilFoundation/lemmy.cafe/
It’s probably something obvious like a missed anchor, but it just doesn’t happen often enough for me to assign mental capacity to this.
Seems like it’s missing ‘restart: always’ in both green and blue. Anubis is lacking those as well though?
Yep, thanks! Pushed the fixed config
You made me took at this now :D Found the problem. Missing
restart: alwaysFWIW I’ve had multiple DigitalOcean VMs die seemingly out of nowhere.
Never had a problem with OVH nor Hetzner.
This was a first for me in ~7 years I’ve been using them. I’ve always found them really good, albeit expensive. And with the ever growing database for lemmy - I’m starting to consider other options.