This message was deleted.

This appears to be an issue that was fixed in Harvester CSI

v0.1.24

(which requires Harvester

v1.6.0

). Issue: harvester/harvester #8066 PR: harvester-csi-driver #59

Hi Webber - just to check that PR seems to be focussed on ungraceful termination of pods and my situation is from graceful termination through the eviction API. Are there improvements for the graceful path as well in that PR?

Your understanding seems incorrect — this applies to nodes in a NotReady state, such as when a node is being drained.