Skip to main content

Troubleshooting

Snapshot restore does not start

Symptoms:

  • PVC from snapshot remains Pending,
  • provisioner events mention source/content errors.

Checks:

  1. kubectl get volumesnapshot -A and verify readyToUse.
  2. kubectl get volumesnapshotclass.
  3. verify dataSource references VolumeSnapshot and correct apiGroup.
  4. inspect controller logs in csi-ipfs namespace.

Snapshot-controller RBAC errors

Symptoms:

  • logs contain forbidden for volumesnapshots, volumesnapshotcontents, or persistentvolumes.

Fix:

  • ensure official RBAC manifest is applied,
  • ensure service account used by snapshot-controller is bound to cluster roles.

Volume publish fails with stale files

Symptoms:

  • mount/publish operations fail due to existing content in destination paths.

Fix:

  • validate node plugin is updated to current logic that clears destination directories before import/export.

E2E cluster issues

Symptoms:

  • tests fail to connect to cluster or use wrong kube context.

Fix:

  • run make -C test/e2e env/up,
  • verify kubectl config current-context is kind-csi-ipfs-e2e,
  • re-run focused suite with GINKGO_FOCUS.

Docs build fails

Run:

yarn --cwd docs install --frozen-lockfile
yarn --cwd docs run build

If broken links are reported, fix sidebar/navbar links and root doc slug.