This message was deleted.

If it’s a single-node Harvester cluster, all the LH volumes are degraded because there is only one replica for each. It’s still usable, but there is no HA. Could you elaborate on what you would like to achieve?

Spread volume across local server drives maybe

You can try accessing the embedded Longhorn UI (here’s how to turn it on) and turning on the

Replica Node Level Soft Anti-Affinity

setting located under the

Setting

page. This will cause Longhorn to schedule replicas on the same node.