Open gavinwilliams opened 2 months ago
Hi @gavinwilliams,
Did you install the NFS CSI driver? That's what installs the nfs.csi.k8s.io
driver.
If you want to use the nfs.csi.akshci.com
driver, you'll need to create a custom StorageClass
(or find the default storage class for this driver), then provide that StorageClass name as part of the Kubernetes agent install UI or via this helm value persistence.storageClassName
Hi @gavinwilliams,
Did you install the NFS CSI driver? That's what installs the
nfs.csi.k8s.io
driver.If you want to use the
nfs.csi.akshci.com
driver, you'll need to create a customStorageClass
(or find the default storage class for this driver), then provide that StorageClass name as part of the Kubernetes agent install UI or via this helm valuepersistence.storageClassName
I did, unfortunately I consistently ran into errors with the pod starting, I believe the error related to having unbound PV claims specifically with the nfs.csi.k8s.io driver.
I'll try what you have suggested once we get our stack back up.
I did, unfortunately I consistently ran into errors with the pod starting, I believe the error related to having unbound PV claims specifically with the nfs.csi.k8s.io driver.
You will see that temporarily as the NFS pod starts up, but once the NFS pod starts, it should bind the PV claim to the tentacle pod.
If the NFS pod is not starting, that could be why you see the unbound PV claim
We're trying to deploy the agent to a K8s cluster behind Azure Stack HCI. We're having issues with PV's, specifically the driver required for NFS on Azure Stack HCI isn't nfs.csi.k8s.io but nfs.csi.akshci.com.
It would be great if we could have this as a configurable option in the values file.