Discourse הפסיק לעבוד על Raspi4

סיכום מבוסס בינה מלאכותית

bekircem reports that Discourse stopped working on their Raspberry Pi 4 after updates, despite previously installing it successfully following a guide linked here. Falco requests diagnostic output, including docker ps -a, ps aux | grep cloudflared, and curl -I localhost. The logs show the app is running but returning 404 errors for /latest, which is unusual. Falco suspects database corruption, possibly due to SD card issues common on Raspberry Pi, even though bekircem confirms no manual database changes were made and only upgraded Discourse. Falco advises restoring from the last known good backup if corruption is confirmed. No plugins are in use, only a theme component. The issue appears to stem from a misconfigured or corrupted database rather than a code or plugin problem.

Hi,

After some updates Discourse stooped working on Raspi4.

I wrote and installed it here before.

I tried many things but I can’t solve the problem. Honestly, I couldn’t find where the error was.

Can you please share the output of

docker ps -a

ps aux | grep cloudflared

docker ps -a

ps aux

What about curl -I localhost ?

Oh that means that the Discourse app isn’t capable of booting.

I just rebuild https://discourse-on-a-pi.falco.dev/ so I know it’s working fine. Do you have any plugins on that install?

tail -n 50 /var/discourse/shared/standalone/log/rails/production.log may also give some clues.

No, i am not using any plugin. I am just using plugin outlet locations theme component.

tail -n 50 /var/discourse/shared/standalone/log/rails/production.log

Processing by ListController#latest as HTML
  Rendered list/list.erb within layouts/application (Duration: 63.6ms | Allocations: 11276)
  Rendered list/list.erb within layouts/application (Duration: 66.0ms | Allocations: 11276)
Started GET "/" for 188.119.17.7 at 2022-08-03 14:48:43 +0000
Processing by ListController#latest as HTML
  Rendered list/list.erb within layouts/application (Duration: 78.8ms | Allocations: 11276)
Started GET "/" for 188.119.17.7 at 2022-08-03 14:48:55 +0000
Processing by ListController#latest as HTML
  Rendered list/list.erb within layouts/application (Duration: 59.2ms | Allocations: 11276)
start
done
start
done
Started GET "/" for 188.119.17.7 at 2022-08-03 14:54:33 +0000
Processing by ListController#latest as HTML
  Rendered list/list.erb within layouts/application (Duration: 51.5ms | Allocations: 11276)
Started GET "/" for 188.119.17.7 at 2022-08-03 14:56:09 +0000
Processing by ListController#latest as HTML
  Rendered list/list.erb within layouts/application (Duration: 52.6ms | Allocations: 11276)
start
done
Started GET "/" for 188.119.17.7 at 2022-08-03 14:57:18 +0000
Processing by ListController#latest as HTML
  Rendered list/list.erb within layouts/application (Duration: 52.1ms | Allocations: 11276)
start
done
start
done
start
done
start
done
start
done
Started HEAD "/" for 172.17.0.1 at 2022-08-03 15:13:18 +0000
Processing by ListController#latest as */*
  Rendered exceptions/not_found.html.erb within layouts/no_ember (Duration: 67.5ms | Allocations: 6453)
Started GET "/" for 188.119.17.7 at 2022-08-03 15:13:43 +0000
Processing by ListController#latest as HTML
  Rendered list/list.erb within layouts/application (Duration: 56.6ms | Allocations: 11272)
Started HEAD "/" for 172.17.0.1 at 2022-08-03 15:14:13 +0000
Processing by ListController#latest as */*
  Rendered exceptions/not_found.html.erb within layouts/no_ember (Duration: 2.0ms | Allocations: 94)
Started HEAD "/" for 172.17.0.1 at 2022-08-03 15:14:35 +0000
Processing by ListController#latest as */*
  Rendered exceptions/not_found.html.erb within layouts/no_ember (Duration: 1.6ms | Allocations: 94)
start
done
start
done

That’s super weird! I’m not aware of cases where the /latest route can cause a 404 :thinking:.

Looks like you ended up in an invalid configuration somehow. Have you messed up with the database?

Routes like https://dev.zettabayt.com/about.json are working just fine, but you have an inconsistency error in your db somewhere.

I’m just upgraded Discourse to the latest version. I did not do anything with the database.

There is a possibility that the database got corrupted because of the Pi SD card usage. In this case you would need to restore from the last working backup.