[SOLVED] How to passthrough a hardware RAID5 volume to a VM

tuxor2

New Member
Apr 19, 2023
6
0
1
Hi,

I'm new on Proxmox but i exalt day after day discovering the solution :)

I have a HP Prolian ML350P, with a P420i hardware RAID Controller managing :
- 2 disks of 450GB in a RAID0 volume
- 4 disks of 3TB in a RAID5 volume (with 1 spare disk)

Proxmox installation well done.
I create a LVM-Thin on my RAID0 volume and all is good on the GUI.

Now, i would like to passthrough all my RAID5 volume in a VM.
I see the disk in my pve console :
code_language.shell:
root@proxmox:~# lsblk -o +MODEL,SERIAL,WWN
NAME                                               MAJ:MIN RM   SIZE RO TYPE MOUNTPOINT MODEL               SERIAL               WWN
.....
sdb                                                  8:16   0   5.5T  0 disk            LOGICAL_VOLUME      001438023C5CBE0      0x600508b1001c8dae7d2d465b32eb8c74
.....

But i'm getting confused when searching the disk info :
code_language.shell:
root@proxmox:~# ls -l /dev/disk/by-id/ | grep sdb
lrwxrwxrwx 1 root root  9 Apr 10 20:35 lvm-pv-uuid-zwvWun-Rtgc-7abb-xUFR-uNeR-pb2W-7xbVpO -> ../../sdb
lrwxrwxrwx 1 root root  9 Apr 10 20:35 scsi-3600508b1001c8dae7d2d465b32eb8c74 -> ../../sdb
lrwxrwxrwx 1 root root  9 Apr 10 20:35 wwn-0x600508b1001c8dae7d2d465b32eb8c74 -> ../../sdb


I would like to point out that I have already had some problems with TrueNAS and that I want to keep the hardware RAID5.


Could you help me please ?

Thanks by advance.
 
Last edited:
Doesn’t my question inspire you?


Please HELP.

/UPDATE profil SET login=Leeloo WHERE login=tuxor2 :D
 
Thank you very much :)
I’m so sorry I didn’t get it the first time...

One last question though: should i take lvm-pv-uuid-zwvWun-Rtgc-7abb-xUFR-uNeR-pb2W-7xbVpO or scsi-3600508b1001c8dae7d2d465b32eb8c74 to passthrough (cf. first post of the thread) ??
Can’t uuid change on reboot ?? Do I have reason to believe that it's better to take the scsi-3600508b1001c8dae7d2d465b32eb8c74 ??
 
Last edited:
Thank you very VERY much :)
For the second question, I try to do some tests and I will update the post
 
As agreed, I come to close the subject with my latest tests.

I disconnected 1 disk on the 3 disks of my raid5 volume to force rebuild.
After few hours, the raid5 volume and datas ares alive.


/!\/!\ However, my fear is confirmed, the UUID changes after rebuild.


In the first post of this thread :
Code:
root@proxmox:~# ls -l /dev/disk/by-id/ | grep sdb
lrwxrwxrwx 1 root root  9 Apr 10 20:35 lvm-pv-uuid-zwvWun-Rtgc-7abb-xUFR-uNeR-pb2W-7xbVpO -> ../../sdb
lrwxrwxrwx 1 root root  9 Apr 10 20:35 scsi-3600508b1001c8dae7d2d465b32eb8c74 -> ../../sdb
lrwxrwxrwx 1 root root  9 Apr 10 20:35 wwn-0x600508b1001c8dae7d2d465b32eb8c74 -> ../../sdb

After rebuild :
Code:
root@proxmox:~# ls -l /dev/disk/by-id/ | grep sdb
lrwxrwxrwx 1 root root  9 May  5 17:13 lvm-pv-uuid-mG67i0-py7Y-Jl58-nx5m-ihn3-RhpB-F1cM54 -> ../../sdb
lrwxrwxrwx 1 root root  9 May  5 17:13 scsi-3600508b1001c8dae7d2d465b32eb8c74 -> ../../sdb
lrwxrwxrwx 1 root root  9 May  5 17:13 wwn-0x600508b1001c8dae7d2d465b32eb8c74 -> ../../sdb



/!\/!\ It's therefore important to use the scsi-* id to passthrough to the vm.
Code:
qm set XXX -scsi1 /dev/disk/by-id/scsi-3600508b1001c8dae7d2d465b32eb8c74



Thanks again.