I am using Vikunja w/ Docker with Traefik in front of it.
I have 60+ docker containers with similar setup. But Vikunja always has a problem, every 12-36 hours I will start getting bad gateway. If I test from localhost and port I can access Vikunja, so I know it’s up. But through Traefik I get nothing. Both Vikunja and Traefik have no log entries for it, and all logs for Vikunja stop as there is nothing reaching it anymore.
Restarting either Traefik or Vikunja fixes it. Vikunja docker is the only one that does this.
Sounds like Traefik might be the problem, if Vikunja is still up. Can you access Vikunja from inside the Traefik container? Even if it’s just with curl or something.
Might be worth checking some basics like memory use and disk space on the server too, if you haven’t already.
(I can’t offer much advice for Traefik. v1 was like magic and easy to configure. v2 was so much hard work and so confusing I eventually switched to Nginx for my reverse proxy needs and found easier.)
Yeah localhost. But if, when it’s broken, you docker exec -it /bin/sh etc into the Traefik container and try to reach Vikunja from there (if curl isn’t available you should be able to install it), you’ll know whether the problem is with the Traefik container not connecting to Vikunja or the Vikunja container not responding to Traefik.