Skip to content

fix(volume): stage a multi-node volume once per node - #542

Draft
QuentinBisson wants to merge 3 commits into
giantswarmfrom
fork/stage-a-volume-once-per-node
Draft

QuentinBisson wants to merge 3 commits into
giantswarmfrom
fork/stage-a-volume-once-per-node

Conversation

@QuentinBisson

Copy link
Copy Markdown

The PR title becomes the release note. Write it for users: what changed for them, not how the code changed.

atelet staged a read-write-many or read-only-many volume once per target, so a CSI driver that tracks its stagings (csi-driver-host-path) refused an actor's second mount of the same volume, its Session directory and the mirrors, with volume … is already staged at …. The CSI spec allows one staging_target_path per volume. atelet now stages such a volume once per node at one path all its targets share, records each target that has it published beside it, and unstages it after the last one, so one actor's unmount still never unstages it under another. A staging an earlier release made per target is unstaged at its own path. csi-driver-nfs advertises no staging and is unaffected.

The ledger row existing-volume-mounted-per-actor-at-sub carries the commit.

  • Tests pass
  • Appropriate changes to documentation are included in the PR

Signed-off-by: QuentinBisson <quentin@giantswarm.io>
Signed-off-by: QuentinBisson <quentin@giantswarm.io>
@QuentinBisson QuentinBisson changed the title A read-write-many volume is staged once per node, so a CSI driver that tracks its stagings accepts a second mount of it fix(volume): stage a multi-node volume once per node Oct 24, 2026
Signed-off-by: QuentinBisson <quentin@giantswarm.io>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant