Closed matlag closed 2 years ago
Maybe you should upgrade to 2.4.2~ynh1, to see if you still have the issue
Sorry, forgot to come back on this: no improvement with the latest version.
I ended up running a full vacuum jpb on the database as documented here:
https://docs-develop.pleroma.social/backend/administration/CLI_tasks/database/
And it worked! I've now run for a few days without issue. So apparently it was more a maintenance issue on my end than the package itself.
Closing!
Describe the bug
Service starts, front-end works, but no backend: front-end shows same timeline since problem started (displays last toot from cache from several days ago). Logs show PostgreSQL connection fails systematically at start.
Context
Steps to reproduce
-Stop service -Start service
Expected behavior
Normal start of service. Front-end continuously updates timeline
Logs
In logs, domain name replaced with
domain.ext
journalctl -f -u pleroma.serviceAnd more logs after, but no more reference to PGSQL.
tail -f /var/log/postsgresql/postgresql-11-main.log
I found some possible tweaks on the config file, so I tried these in config.exs (DB password replaced with
**********
, but I can manually connect to the DB from a psql terminal copy-pasting the password from the config file, so that's not the DB credentials.But I got the same result, so I reverted the change.
I have no idea where to go from here.