Problem with volume (partition lost?)

[telatin]

A volume previously attached to a VM looks like lost its only partition.

I’m member of the IFRQMRA team, and the issue belongs to the Bham datacenter.

The volume, labeled “ngs” (attached as /dev/vdc to the VM labeled “tarp”), was mounted as /dev/vdc1. Now in my device list I only see the drive (/dev/vdc) but I’m no longer able to see the content.

This happened some time ago, but before re-formatting and re-uploading the latest backup, I wanted to know if you can see what happened to the drive, just in case I did something wrong to be avoided in the future.

Cheers, have a good day

Andrea

[mattbull]

This is a weird one, we’ve never seen this behaviour before.

Before trying anything else, please could you reboot this VM and then check the state of the volume again?

You can reboot from the command line:

sudo reboot

or use https://bryn.climb.ac.uk to issue a reboot.

[telatin]

Not much luck with this.
Just hoped you could see something wrong from the logs.

NO problem, I’ll reformat and reload the content.

Cheers!
Andrea

[mattbull]

Our logging only shows what’s happening outside volumes, nothing internal to the volume/instance.

As this is entirely unexpected behaviour, it would be good to know what’s going on here.

Please hold-off reformatting this disk for a moment, I’ll send you a PM to move this investigation on further.

[mattbull]

For the sake of new readers of this post - we recovered this volume and all of the data within it.

@telatin took exactly the right approach and let us know about the problem before taking the nuclear option and restoring data from a backup. It was also great to know that @telatin had a restorable backup of the data in this volume, just in case!

In the end, the VM that this volume was attached to needed to be rebooted (a.k.a power-cycled) using the reboot function in https://bryn.climb.ac.uk, a simple OS reboot was not enough to force the hypervisor to re-check the volumes that this VM was attached to.