Restoring GCP Regional pv error

[christianlafleur]

Description

Restore for regional storage class on GCP fail with this error : Invalid value for field 'resource.replicaZones[0]': ''. Exactly 2 replica zones must be set for regional disks

Standard storage class restore is working.

The output of the following commands will help us better understand what's going on:

ark create restore test2 --from-backup full01-20190204180049 --include-namespaces chris-test
Restore request "test2" submitted successfully.

ark describe restore test2
Name: test2
Namespace: heptio-ark
Labels:
Annotations:

Backup: full01-20190204180049

Namespaces:
Included: chris-test
Excluded:

Resources:
Included: persistentvolumes
Excluded: nodes, events, events.events.k8s.io, backups.ark.heptio.com, restores.ark.heptio.com
Cluster-scoped: auto

Namespace mappings:

Label selector:

Restore PVs: auto

Phase: Completed

Validation errors:

Warnings:
Ark:
Cluster:
Namespaces:

Errors:
Ark:
Cluster: error executing PVAction for /tmp/878522100/resources/persistentvolumes/cluster/pvc-0c5099ee-289f-11e9-bbb9-42010aa20034.json: rpc error: code = Unknown desc = googleapi: Error
400: Invalid value for field 'resource.replicaZones[0]': ''. Exactly 2 replica zones must be set for regional disks., invalid
Namespaces:

Anything else you would like to add:

kubectl get pv pvc-5d8cb4c0-28bf-11e9-bbb9-42010aa20034 --show-labels

NAME CAPACITY ACCESS MODES RECLAIM POLICY STATUS CLAIM STORAGECLASS REASON AGE LABELS
pvc-5d8cb4c0-28bf-11e9-bbb9-42010aa20034 10Gi RWO Delete Bound chris-test/pvc-regional regional 56s failure-domain.beta.kubernetes.io/region=northamerica-northeast1,failure-domain.beta.kubernetes.io/zone=northamerica-northeast1-a__northamerica-northeast1-b

Environment:

• Ark version (use ark version): 0.10.1
• Kubernetes version (use kubectl version): v1.10.7
• Kubernetes installer & version: "v1.10.11-gke.1"
• Cloud provider or hardware configuration: GKE

[nrb]

Thanks for the bug report! Would you mind also providing the output of kubectl get pv kubectl get pv pvc-5d8cb4c0-28bf-11e9-bbb9-42010aa20034 -o yaml?

[christianlafleur]

apiVersion: v1
kind: PersistentVolume
metadata:
annotations:
kubernetes.io/createdby: gce-pd-dynamic-provisioner
pv.kubernetes.io/bound-by-controller: "yes"
pv.kubernetes.io/provisioned-by: kubernetes.io/gce-pd
creationTimestamp: 2019-02-04T20:56:42Z
finalizers:

• kubernetes.io/pv-protection
  labels:
  failure-domain.beta.kubernetes.io/region: northamerica-northeast1
  failure-domain.beta.kubernetes.io/zone: northamerica-northeast1-a__northamerica-northeast1-b
  name: pvc-5d8cb4c0-28bf-11e9-bbb9-42010aa20034
  resourceVersion: "146304170"
  selfLink: /api/v1/persistentvolumes/pvc-5d8cb4c0-28bf-11e9-bbb9-42010aa20034
  uid: 5fc9b00d-28bf-11e9-bbb9-42010aa20034
  spec:
  accessModes:
• ReadWriteOnce
  capacity:
  storage: 10Gi
  claimRef:
  apiVersion: v1
  kind: PersistentVolumeClaim
  name: pvc-regional
  namespace: chris-test
  resourceVersion: "146304136"
  uid: 5d8cb4c0-28bf-11e9-bbb9-42010aa20034
  gcePersistentDisk:
  fsType: ext4
  pdName: gke-devl01-671c7607-dy-pvc-5d8cb4c0-28bf-11e9-bbb9-42010aa20034
  persistentVolumeReclaimPolicy: Delete
  storageClassName: regional
  status:
  phase: Bound

[nrb]

Thanks for that, @christianlafleur. I've managed to replicate the issue and have a fixed that I've tested locally. I'm going to clean up the code, add some tests, then push up a PR.