I had some open threads but hit a limit on my codex subscription, and had to rotate my credentials. Once I rotated the credentials and restarted the k3's all existing threads were basically bricked with the following error:
Failed to start the runtime: failed to resume suspended sandbox: centaur-centaur-sandbox-slack-<redacted>
I manually update the postgres records to mark the containers closed (I believe "DONE") but that didn't help so I ended up clearing out the pods manually.
Setup
I had some open threads but hit a limit on my codex subscription, and had to rotate my credentials. Once I rotated the credentials and restarted the k3's all existing threads were basically bricked with the following error:
I manually update the postgres records to mark the containers closed (I believe "DONE") but that didn't help so I ended up clearing out the pods manually.
Setup