Skip to content

ateapi: skip deleting external volumes when Terminate or volume detach fails - #3112

Open
Abhay Joshi (abhayjoshi201) wants to merge 1 commit into
agent-substrate:mainfrom
abhayjoshi201:fix/delete-actor-skip-volumes-on-terminate-err
Open

Abhay Joshi (abhayjoshi201) wants to merge 1 commit into
agent-substrate:mainfrom
abhayjoshi201:fix/delete-actor-skip-volumes-on-terminate-err

Conversation

@abhayjoshi201

Copy link
Copy Markdown

DeleteActor called ensureVolumesDeleted even when ensureAteletTerminated or ensureVolumesDetachedForDelete returned an error. When a CSI driver deletes a still-published volume and returns NotFound from NodeUnpublishVolume on the retried Terminate, atelet skips unmounting and resetActorDirs fails with EBUSY on the mounted volume directory.

Skip ensureVolumesDeleted whenever ensureAteletTerminated or ensureVolumesDetachedForDelete fails, matching ensureWorkerReleased, so external volumes are deleted only after node termination and volume detachment succeed.

Fixes #2971

AI use: AI-assisted; I reviewed the diff and verified the unit and functional test suites locally.

  • Tests pass
  • Docs updated

…h fails

DeleteActor called ensureVolumesDeleted even when ensureAteletTerminated
or ensureVolumesDetachedForDelete returned an error. When a CSI driver
deletes a still-published volume and returns NotFound from
NodeUnpublishVolume on the retried Terminate, atelet skips unmounting
and resetActorDirs fails with EBUSY on the mounted volume directory.

Skip ensureVolumesDeleted whenever ensureAteletTerminated or
ensureVolumesDetachedForDelete fails, matching ensureWorkerReleased, so
external volumes are deleted only after node termination and volume
detachment succeed.

Signed-off-by: Abhay Joshi <abhayjoshi@google.com>
@abhayjoshi201
Abhay Joshi (abhayjoshi201) force-pushed the fix/delete-actor-skip-volumes-on-terminate-err branch from 3cb8cdc to ccd6750 Compare November 11, 2026 11:12

This branch has not been deployed

No deployments
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

ateapi, atelet: retried actor delete gets EBUSY when volumes were deleted after a failed Terminate

2 participants