David J. M. Karlsen

> @davidkarlsen Currently we support ubuntu based systems only, if possible, try to change the host OS to ubuntu. Also, is the issue reproducible everytime you reboot the machine? not...

> @davidkarlsen - I wonder if this may be related to [access.redhat.com/solutions/2175491](https://access.redhat.com/solutions/2175491) Hmm, how can I figure out if multipath is in use? But I doubt it as we're on:...

> How is the reboot of the machine triggered? Does it clear all the volumes? Can you try to capture the lvm config file after volumes are provisioned and post...

could this perhaps be a se-linux issue? I assume the lvm command are executed into the host-system from one of the openebs containers, and that might not have enough privilegies?...

This should be of interest: ``` docker run --rm -ti --entrypoint /bin/sh k8s.gcr.io/sig-storage/csi-node-driver-registrar:v2.3.0 docker: Error response from daemon: OCI runtime create failed: container_linux.go:380: starting container process caused: exec: "/bin/sh": stat...

> @davidkarlsen can you execute `pvscan --cache` (and lvscan) command before doing node reboot. I am wondering whether node cache is not updated before reboot, if not, we might lose...

Ok, so I've tried creating a bunch of volumes and rebooting machines. Two changes: * I noticed the previous openebs install was not 100%, because in some upgrade the storageCapacity...

> @davidkarlsen Could you reopen this issue? reopened

@npu21 what OS do you run on the nodes?

Why would this be needed? Doesn't the service run normally, and users can run LVM commands without LVM getting corrupt?