The rancher UI somettimes hangs on "loading" forever when listing pods or other resources. (kubectl get pods works). I had this before and it seemed gone after deleting 900 Evicted pods. Now it's back and after some digging it seems that the cattle-cluster-agent has an issue. It's emitting this every 30 seconds: time="2025-10-28T103303Z" level=error msg="Error in Store.Add for type _v1_Pod: transaction: begin tx: database is locked (5) (SQLITE_BUSY)" After restarting the pods it's working again but I'd like to know why. Does cattle-cluster-agent use SQLite internally?

On recent releases, the UI has a sqlite based cache layer between UI and apiserver for sorting and filtering. This is covered in the docs.

That seems to be the issue yeah thanks. I'll take a look on the gc treshold. It happens on the local cluster where rancher itself is running (k3s with mysql)