lvm-thin on SSD stalls/fails clone or create

[sholinaty]

steps:
1. installed a new 1tb SSD into server
2. configured a new thinpool on the web UI
3. hit Create VM... set the target to my new thinpool
- I see the "Logical olume "vm-106-disk-0" created message
- the create never actually finishes (left it running for 24hrs once, just in case)

or 3. right click on a template... Clone... FullClone, target = my new thinpool
-- disk is created, transfer starts
- transfer stalls out every time at exactly 7.4gb or 20gb
also tested it cloning a different non-template VM, dies at Exactly the same point.

Any suggestions on where to start?

 

[bbgeek17]

dmesg
journalctl -b

---

Blockbridge: Ultra low latency all-NVME shared storage for Proxmox - https://www.blockbridge.com/proxmox

 

[sholinaty]

from"syslog
Apr 14 15:18:33 proxmox1 kernel: [1055476.174774] device-mapper: thin: dm_thin_get_highest_mapped_block returned -5

dmesg is the same:
[1055535.600523] device-mapper: thin: dm_thin_get_highest_mapped_block returned -5

[1055535.601022] device-mapper: thin: dm_thin_get_highest_mapped_block returned -5

[1055535.601529] device-mapper: thin: dm_thin_get_highest_mapped_block returned -5

[1055535.832456] device-mapper: thin: dm_thin_get_highest_mapped_block returned -5

[1055535.834789] device-mapper: thin: dm_thin_get_highest_mapped_block returned -5

[1055535.835576] device-mapper: thin: dm_thin_get_highest_mapped_block returned -5

[1055535.836356] device-mapper: thin: dm_thin_get_highest_mapped_block returned -5

[1055546.225200] device-mapper: thin: dm_thin_get_highest_mapped_block returned -5

[1055546.226326] device-mapper: thin: dm_thin_get_highest_mapped_block returned -5

[1055546.226804] device-mapper: thin: dm_thin_get_highest_mapped_block returned -5

[1055546.227311] device-mapper: thin: dm_thin_get_highest_mapped_block returned -5

[1055546.473502] device-mapper: thin: dm_thin_get_highest_mapped_block returned -5

[1055546.474825] device-mapper: thin: dm_thin_get_highest_mapped_block returned -5

[1055546.475488] device-mapper: thin: dm_thin_get_highest_mapped_block returned -5


[1055546.476199] device-mapper: thin: dm_thin_get_highest_mapped_block returned -5

 

[bbgeek17]

A google search for that message, albeit with few results, indicates an lvm metadata corruption. I would also consider your new disk suspect as well. I would try to read/write entire disk with "dd" and check its SMART stats.
If hardware checks out ok, I would erase and rebuild the LVM on it.

---

Blockbridge: Ultra low latency all-NVME shared storage for Proxmox - https://www.blockbridge.com/proxmox

 

[sholinaty]

that would likely make sense - as shown by

the disk labelled thinpool is my SSD... data was an old 1tb spinning disk.

I'm wondering if I just need to manually specify the metadata size when provisioning the thinpool.

I have tried wiping and reinitializing the disk a few times now.

 

[sholinaty]

still stuck on this.
I was able to extend the metadata, and everything "Seems" right, bu I am still getting the syslog errors..

proxmox1 kernel: [1396195.867975] device-mapper: thin: dm_thin_get_highest_mapped_block returned -5

wiped the disk, same errors.

 

[Docop2]

i always get problem with adding disk and the solution quite easy is to just boot over a partition manager soft and create lvm there. Then reboot and hop all detect in prox.

 

[sholinaty]

i always get problem with adding disk and the solution quite easy is to just boot over a partition manager soft and create lvm there. Then reboot and hop all detect in prox.

problem is, having to boot into a partition manager causes downtime. the whole point of adding extra drives is to help ensure critical services remain Up.