VMware vSphere

 View Only

 vSphere 7.0.3 can't add datastore after raid 5 rebuild

Thomas Bach's profile image
Thomas Bach posted Jul 14, 2025 09:39 PM

One drive has collaplsed, another one failed while I was in a process of rebuilding first one, and it did bring whole virtual drive group down and made it inaccessible. I still managed to swap bad drives and finished rebuilding it, bringing raid5 online, but old datastore still inaccessible, options like delete/unmount/mount are grayed out, and when I'm trying to Add Storage as new one - during partition configuration wants to wipeout layout and data. Is there any way to add old raid as new datastore without losing all the data?

Hypervisor: VMware ESXi, 7.0.3, 23307199, I'm ready to run esxcli commands for more info. 

Octopus's profile image
Octopus

Don't use Add Storage from the UI — it will wipe the partition table.
Resignature is non-destructive and creates a new datastore UUID but preserves contents.

Try, maybe like this, if possible:

SSH into the host if not already:
ssh root@<esxi-host-ip>

List all detected VMFS volumes:
esxcli storage vmfs snapshot list

Try resignature (safe, non-destructive):
esxcli storage vmfs snapshot resignature -l "<datastore-name>"

Or if the name isn't clear:
esxcli storage vmfs snapshot resignature -u <UUID>

try mounting it manually:
vmkfstools -V  # rescan for VMFS
esxcfg-volume -l  # list volumes
esxcfg-volume -M <volume-label or UUID>  # persistent mount

Or try force mount (non-persistent, test only):
esxcfg-volume -m <volume-label or UUID>

Thomas Bach's profile image
Thomas Bach

Coming in kinda late for reply, I did login via ssh and checked 
esxcli storage vmfs snapshot list and resignature 
Sadly it downed on me kind of late, that I can check hexdump 
hexdump -C /dev/disks/deviceid | less 
And header was 00000000 
Raid was gone, losses were cut, nothing really critical was lost, let this thread be a reminder to properly maintain your drives and thanks Octopus for your input.