Is there a way to create a CephFS PVC without quota or remove the quota from the subvolume?

[mmeier86]

Hello,

I'm currently migrating from a cephadm cluster on Nomad to a Rook cluster in k8s, and I seem to have hit a (slight) snag: In my baremetal cluster, I've got a CephFS subvolume which is allowed to consume the entirety of the pool it's running on, with no quotas set.

I'm now trying to figure out how to get the same setup in Kubernetes, a CephFS PVC that is mountable to Pods like any other PVC, but doesn't impose quotas on the underlying CephFS subvolume. Is that possible somehow? I've already tried removing the resources.requests.storage key from the PVC manifest entirely and also set it to zero, but neither of those are allowed by Kubernetes.

Does anyone have an idea? E.g. is there a way I could create the subvolume in Ceph and then setup a PersistentVolume(Claim) that uses it and just ignores the capacity setting?

[Madhu-1]

@mmeier86 i think you can do couple of things

• you can try static PVC creation https://github.com/ceph/ceph-csi/blob/devel/docs/static-pvc.md#cephfs-static-pvc?
• Create a pvc with the size of 1 Gb but you can remove/reset the subvolume quota in backend using ceph fs command.

[mmeier86]

Thank you very much Madhu! It looks like this does exactly what I needed. I've created a static PVC following those instructions and even though I gave it a capacity of 1Gi, the Ceph info output still shows an infinite quota.