Skip to content

Volumes: console walkthrough

The same operations as the step-by-step guide, done in the console.

List

WebHosting > Storage > Volumes lists the persistent volume claims in the selected project.

Read the Status column: Bound means a disk is attached to the claim. A claim sitting at Pending has not been given one yet - open it and read the events at the bottom, which name the reason.

Unlike the other pages in this section, Volumes is the console's own storage page rather than one this platform adds, so it offers the full set of actions rather than a guided subset.

Create

Create PersistentVolumeClaim opens the console's form.

  • Storage class - decides what kind of disk you get. See Choose a storage class for which to pick; the short version is a block class for one workload, a file class for several.
  • Name - how workloads in the project refer to it.
  • Access mode - ReadWriteOnce for one workload, ReadWriteMany for several. The options offered depend on the storage class.
  • Size - can be grown later, never shrunk.

Access mode is the decision to get right

A ReadWriteOnce volume can be mounted by one pod at a time. Give one to a Deployment with two replicas and the second pod stays Pending indefinitely, with an event saying the volume is attached elsewhere. If more than one pod needs the same data, choose a file storage class and ReadWriteMany at creation - it cannot be changed afterwards.

Change

Open the claim and use Actions > Expand PVC to grow it. The new size takes effect when the pod next starts, or immediately if the storage class supports expanding a mounted volume.

Shrinking is not possible, and neither the storage class nor the access mode can be changed after creation.

Delete

Use Actions > Delete PersistentVolumeClaim on the claim, or the kebab menu in the list.

This usually destroys the data

The storage classes on this platform are set to delete the disk with the claim. There is no undo and no snapshot taken on your behalf.

A claim mounted by a running pod stays Terminating until that pod is gone. Scale the workload down rather than forcing the deletion - the protection is what stops a running application losing its disk underneath it.

Where the console differs from the CLI

Console oc
Access modes offered Filtered to what the class supports You can request one it cannot provide
Expanding Actions > Expand PVC oc patch the storage request
Why a claim is Pending Events shown on the claim page oc describe pvc
Reclaim policy Not shown Read it off the storage class