Hi, this server has been fixed and is now in a READY state, ready for normal use.
Status Explanation:
The upgrade timing coincided with the K3s initialization sequence, causing the internal RBAC bootstrap settings to become corrupted after the upgrade, preventing the service from recovering. Specifically, the cloud-controller-manager failed to start, causing the Kubernetes API server to enter a crash loop (with nearly 3,000 restart attempts). This is why you saw K3s as offline in Zeabur, and why automatic reinstalls were failing.
Fix Details:
We SSH'd directly into the server to rebuild K3s (as the RBAC corruption could not be fixed through standard methods), synced the new TLS certificates back to the Zeabur dashboard, and redeployed the Zeabur system components. We confirmed that no server data was lost (your previously deployed bally-doc-erp service had not successfully started before the upgrade, so no running data was affected).
Next Steps:
Please go to the Dashboard, find the bally-doc-erp service, and manually trigger a Deploy. The server is ready.
We will add safeguards to the backend to prevent this upgrade timing issue from happening again. Please continue to report any further issues.
Forum