Hi,
I’m encountering a critical issue with my Longhorn setup, and I would greatly appreciate any guidance or suggestions.
I’m running a Kubernetes cluster on DigitalOcean Managed Kubernetes (DOKS), and my storage backend is configured with Longhorn. After a recent forced Kubernetes upgrade, I found that all my Longhorn volumes are in a Faulted state.
I can still see all the volumes listed in the Longhorn UI with their metadata intact.
The ownerID of the volumes has been updated to the new node names after the upgrade.
Attempts to salvage the volumes have failed with errors like:
“disk with UUID on node is unschedulable for replica .”
DigitalOcean Support confirmed that DOKS nodes have ephemeral storage, and any node rebuild (as happens during upgrades) would clear local storage.
Is there any way to verify if the underlying .img files for these volumes still exist, even after the upgrade?
Any advice, suggestions, or guidance would be deeply appreciated.