WARNING: Not using lvmetad because config setting use_lvmetad=0

[crazy-to-bike]

Hi,

I updated Proxmox 5 just now and a new kernel was installed. I get warnings about lvm while updating the grub config:

WARNING: Not using lvmetad because config setting use_lvmetad=0
WARNING: To avoid corruption, rescan devices to make changes visible (pvscan --cache)

pvscan delivers the same warnings

pvscan
WARNING: Not using lvmetad because config setting use_lvmetad=0.
WARNING: To avoid corruption, rescan devices to make changes visible (pvscan --cache).
PV /dev/sdb1 VG vmdata lvm2 [7,28 TiB / 169,77 GiB free]
PV /dev/sda3 VG pve lvm2 [49,75 GiB / 6,07 GiB free]
Total: 2 [7,32 TiB] / in use: 2 [7,32 TiB] / in no VG: 0 [0 ]

Google search is finding a bug anounced at red hat from this year, the other stuff is older:
https://bugzilla.redhat.com/show_bug.cgi?id=1431158
What's going wrong?
How can I fix it?

Edit:
Here is also some information
https://unix.stackexchange.com/ques...s-running-but-disabled-restart-lvmetad-before
https://unix.stackexchange.com/ques...s-running-but-disabled-restart-lvmetad-before
but I still don't know if I should do anything or if i really can ignore the messages.

regards

 

[GadgetPig]

Hello,

I encountered the same issue awhile back. It sounds scary, but actually harmless. If I interpret correctly, this happens because "pvscan --cache" can't run while the service "lvm2-lvmetad" is running.

Here's what I did to make sure apt-get update/dist-upgrade went smooth:

1. #service lvm2-lvmetad stop (this stops the service so pvscan can run)
2. #apt-get update && apt-get dist-upgrade
3. #pvscan --cache
4. #update-grub (prolly not needed but I ran it anyways)
5. #service lvm2-lvmetad start

You may need to reboot the server. But after the reboot, no more more "lvmetad" issues happened during upgrades

 

[crazy-to-bike]

Hi,

thanks for your answer.

Can I stop lvm2-lvmetad while the server is in use and VM are running?

Thinking about the message

WARNING: Not using lvmetad because config setting use_lvmetad=0

I think about the value of use_lvmetad=0.

In /etc/lvm/lvm.conf

use_lvmetad=0

Shouldn't I set it to use_lvmetad=1 ?

 

[GadgetPig]

"Can I stop lvm2-lvmetad while the server is in use and VM are running?"
In my case I stopped lvm2-lvmetad while2 VM's were running (Windows and ClearOS VM) and I didn't notice any issues during the updates. This is on a standalone (non cluster) server with pve-nosubscription repository.

I can't recall making changes to the lvm.conf file, but I confirmed all my updated servers have lvm2-lvmetad service running. If you keep getting lvm2-lvmetad errors during updates, I guess you can set it to "use_lvmetad=1".

I wish I could help more, but here's what I know about lvm2-lvmetad service:

https://access.redhat.com/documenta...me_Manager_Administration/metadatadaemon.html

If you have important VMs, I always recommend backing up VM data, as well as a full ProxMox VM backup, before doing updates to your ProxMox server. Good Luck!

 

[GadgetPig]

@crazy-to-bike, I took a look at my current lvmetad settings on standalone server, this is what I found:

from /etc/lvm/lvm.conf

use_lvmetad = 0

 

[crazy-to-bike]

thank you for your help and sorry for the late response.

A simple reboot seems having solved the problem.