May have ruined data of one of my nodes...

[TheePorkchopExpress]

Some context before I jump right in the issue.. (also am learning as you can tell by the below, so sorry if there is an easy solution - fingers crossed - or if I am leaving out very important information) Also, no I don't have back ups... I have drives, etc.. just haven't done it yet..

1. I have 2 servers. Server1 and Server2.
2. Those 2 servers have been running proxmox as 2 nodes for about a year.
3. Server2 was purchased with 16 900gb SAS drives
4. Boot disk was in Raid1 with two of those drives
5. "data" was in raid 6 with the remaining 14.
6. 4 or 5 of those 14 were "foreign" but because I wanted to just get started I left those alone and worked with the remaining drives
7. Built some docker containers, and put a lot of effort into my Plex library.
8. I decided I wanted to fix that foreign drives..
9. I cleared those foreign drives in server2's bios
10. When I did that it deleted my VDs, but maybe not the data? No prompt said it would (I did NOT initialize any disks in the process of rebuilding the RAID1 boot and RAID6 data drives).
11. "Successfully" from the servers perspective recreated the VDs arrays, with the previously foreign drives.
12. I figured I would have to do something with proxmox. But wasn't sure. Proxmox was still installed so felt pretty good about things..
13. Then I logged in and saw error messages like "TASK ERROR: no such logical volume data2.2/data2.2" (see screenshot 2 for other errors)
14. In screenshot1 there is the drives, but I cannot create LVM-Thin, Volume group or anything with it. All have a prompt in the UI that says "No unused disks."
15. All data was previously on data2.2, pve2.

Just to be clear (this feels complicated to me) I had my drives working, deleted the VD disk on the server by accident but was able to rebuild those with the same raid array (RAID6) but with a few extra disks this time - how do I "point" my VMs/etc to these drives? How do I reclaim that data and set up? Am I screwed?

 

[danielb]

You should recreate the RAID array with the exact configuration as it was previously (same PD, same order, same chunk size etc.)

 

[TheePorkchopExpress]

You should recreate the RAID array with the exact configuration as it was previously (same PD, same order, same chunk size etc.)

Darn, so I need to remove the physical disks that weren't there before, then reattempt? Then once that is working add the other drives? (all the rest of the config was the same )

 

[danielb]

Depends on your RAID card if you'll be able to expand the RAID6 array with more PD. But at least, you might be able to get your data available, back them up, wipe and recreate the array on all the disks, and restore

 

[TheePorkchopExpress]

Depends on your RAID card if you'll be able to expand the RAID6 array with more PD. But at least, you might be able to get your data available, back them up, wipe and recreate the array on all the disks, and restore

Yeah I was able to succefully add all the drives to my RAID6 array, but in the previous array config before I deleted my VD somehow, there were less drives. Both times it was RAID6, all the other configs for the array were the same, just this time I added some more drives.

So I should

1. remove any "new" drives from the array, i.e. so that the number of drives before the accident and after the accident are the same?
2. then the data should be available? and I can re-add in Proxmox? Hopefuly in UI?
3. lastly add more disks to the array and bring them into proxmox?

 

[danielb]

You should delete the actual RAID configuration (delete the VD again), then re-create the VD exactly as it was previously (same set of physical drives, in the same order, with the same chunk size etc. Everything must be exactly the same). Be careful not to initialize the VD when recreating it. If you manage to get the VD back in the same state, it should be available in Proxmox GUI again. Then I'd suggest making backups, wipe the VD, and recreate it again with all the physical disks