How to unmount disks in VSOM 7.9?

[mstrz]

I need to remove all the 2TB disks, media1-5. What's the process to remove them completely from VSOM?

When I run df -h, this is the output:

Filesystem Size Used Avail Use% Mounted on
/dev/sda2 7.9G 5.5G 2.1G 73% /
tmpfs 5.4G 524K 5.4G 1% /dev/shm
/dev/sda8 53G 8.9G 42G 18% /mysql/data
/dev/sda6 7.9G 4.7G 2.9G 62% /usr/BWhttpd
/dev/sda4 32G 1.4G 29G 5% /var
/dev/sdc1 2.0T 34M 2.0T 1% /media1
/dev/sdd1 2.0T 61M 2.0T 1% /media2
/dev/sde1 2.0T 47M 2.0T 1% /media3
/dev/sdf1 2.0T 61M 2.0T 1% /media4
/dev/sdg1 2.0T 33M 2.0T 1% /media5
/dev/sdb1 8.0T 6.1T 2.0T 76% /media6

[Scott Olsen]

Well... Is this server also running VSMS?  If not, this is strictly an OS administration thing as far as I'm aware.

If it's still running VSMS my approach would be to unselect all the media repos under the advanced settings for the VSMS role and then remove the VSMS role completely first.

If VSMS is no longer part of the server in any way, I'd simply comment out the mounts in /etc/fstab and reboot the server.  Assuming this is a VM (which it hints at being), the step after that would be removing those additional block devices (hard disks) that provided SDC-G.

[mstrz]

Yes, this is on a VSMS server (VM), but I'm not trying to remove the VSMS role. I'm only removing some of the mounted disks.

There are also some shelved recordings that for some reason never got groomed. I'm not sure if it's safe to just delete those?

[Scott Olsen · Accepted solution]

Sorry, I just re-read your post and missed the fact that you wanted to keep /media6.

In that case, I'd make sure media repos 1-5 are *de-selected* in the VSMS configuration, then comment out those specific mounts and reboot to test and ensure you still have /media6 available and functional.

If you can successfully get to this point, the final challenge will be identifying which block devices attached at the VM host level represented those hard disks (SDC-G).  I'm not 100% sure on how to do that myself (in a VMWare env.).

I'm assuming you *might* be able to look at the actual consumed storage on the host datastore (depending on how the disks were provisioned).  It appears only the disk representing /DEV/SDB has any considerable usage.  That might be a way to narrow it down?

Cheers!

[mstrz]

Thanks! I was unmounting the disks and they kept coming back, I didn't know to check fstab.