Nocodb stop responding after about every 2 days
-
After 36-48 hours, my nocodb instance gets down.
I usually reboot Cloudron (there is only another unused 8n8 app on it) and then I'm good for anew period.
I don't know how to tackle this pbl.
The server is a vps lite from Infomaniak with 4 cpu, 4 Gb of ram and 80 Gb of disk. I already migrated from a public cloud infomaniak slightly less powerfull (2 cpu) because of the same problem, thinking it was the lack of power. But it didn't change. Even got worse (more frequent) !
There is not so much data in this nocodb. 6 databases very simple. Some of them are just one table. Most bases are used as backend for small nuxtjs website. Most api access are heavily cached until invalidated on the nuxt sites so I guess there is very little charge on the server.
Looking at the logs, I see frequently error like this: "Nov 10 21:22:39 [Nest] 1 - 11/10/2025, 8:22:39 PM LOG [BaseModelSqlv2] Error: Formula error: Knex: Timeout acquiring a connection. The pool is probably full. Are you missing a .transacting(trx) call?" that I don't understand. It doesn't seem to prevent the app to be functioning.
Looking at the graph I can see that the ram consumption is slowly rising, starting at 0,4 gb after a reboot and rising to 1 gb where it usually doesn't last long before the app become irresponsive. Even if my memory limit is set to 4gb or ram and 100 % of cpu.
Cloudron itself is always working fine and responsive.
I'd be very grateful if someone could guide me or even have a look ! Thank you. -
Just a wild guess, but sounds like maybe something in nocodb is not freeing up database connections and eventually no more new connection can be established. Restarting the app would free up those handles.
Have you searched upstream in the nocodb issue tracker already?
