Site icon techbeatly

How to Fix a Failed PV in OpenShift Cluster

Image : https://premierevanlines.com

There are several cases a PV (PersistentVolume) appear as Failed in OpenShift or Kubernetes cluster. Once of the reason is wrong ClaimRef by which PV will mark as claimed but actually not in use. Let’s see one example.

# oc get pv |grep lv-1g-110
lv-1g-110 1Gi RWO,ROX,RWX Recycle Failed newproject/postgresql-test-dep 26d

From above snippet we can understand that, the PV lv-1g-110 is claimed by a PVC (PersistentVolumeClaim) named postgresql-test-dep and this is under project newproject. So let us check that PVC details.

# oc get pvc -n newproject | grep postgresql-test-dep
# oc get pvc -n newproject | grep lv-1g-110
#

Aah ! now we got confused; there is no such PVC called postgresql-test-dep or anything with lv-1g-110. Means, lv-1g-110 is stuck with some wrong claim. Let us see what is holding inside PV config.

# oc describe pv lv-1g-110 |grep Claim
Claim: newproject/postgresql-test-de

No wonder, we can see the PV got a ClaimRef for a PVC which doesn’t exist at all ! Let’s remove this.

Method 1 : Using oc edit

oc edit and remove those ClaimRef lines.

# oc edit pv lv-1g-110

Then remove below lines and save it.

claimRef:
apiVersion: v1
kind: PersistentVolumeClaim
name: postgresql-test-de
namespace: newproject
resourceVersion: "569221692"
uid: 44aa4a1a-e3fd-11e8-83ad-65efe91e200

Method 2 : Using oc patch

# oc patch pv/lv-1g-110 --type json -p '[{ "op": "remove", "path": "/spec/claimRef" }]'

Now we check the pv status.

# oc get pv |grep lv-5g-293
lv-1g-110 1Gi RWO,ROX,RWX Recycle Available 8

That’s all; our pv is Available for next claim.

Exit mobile version