NVME for Metadata Special Device

[janos66]

Hi,

I configured two 512G SAMSUNG MZVPV512HDGL as special device in a mirror for my mail zpool, they are old but I thought it's a mirror what could go wrong?

By chance I checked the boot log shortly afterwards and found this entry

Code:

Feb 18 05:20:38 truenas.vmware.intern  smartd[5907]: Device: /dev/nvme0n1, number of Error Log entries increased from 3225 to 3226
Feb 18 05:20:39 truenas.vmware.intern  smartd[5907]: Device: /dev/nvme1n1, number of Error Log entries increased from 4108 to 4109

Maybe I interpret this incorrectly, but now I prefer to look for new and optimally suitable NVMEs for the sepcial device. The zpool has 100TB, what would you recommend

 

[winnielinnie]

I think that's due to the firmware + SMART not understanding the NVMe properly. You're better off disabling/ignoring SMART for those drives and rely on nvme tools instead.

I get the same thing on two different Linux computers with NVMe drives. Two of mine are also Samsungs.

If you really want to test it out, see if you can increase the "error log entries" each time you restart the server / VM. My hunch tells me it'll increase it each time on queue.

Related discussion on the Proxmox forums: https://forum.proxmox.com/threads/pve-7-1-new-smart-error-on-every-hard-disk-at-every-reboot.101063/

 

[janos66]

Thank you very much, that saves my money and I will take a closer look at it again