HDD Sleep not working

[Nero_XIII]

Hi everyone,

i do need some help and hopefully someone knows where the problem is. My setup:
Proxmox installation on NVME, 2 Physical HDD's (besides some others) attached as Data Storage for the TrueNAS Core which is running in a VM.
My problem is that those drives are not needed all the time and should spin down / go to sleep when not used.

I already googled and searched for this and tried some things. I configured the /etc/lvm/lvm.conf and added this line:

global_filter = [ "r|/dev/zd.*|", "r|/dev/mapper/pve-.*|", "r|/dev/mapper/.*-(vm|base)--[0-9]+--disk--[0-9]+|", "r|/dev/sda*|", "r|/dev/sdb*|", "r|/dev/sdc*|", "r|/dev/sdd*|" ]

But the drives still get some R/W's. I observed them for some minutes by using

blktrace -d /dev/sda -o - | blkparse -i -"

and i can see there are reads and writes every few minutes. Problem is i cannot figure out what process is doing this. This is just a little part of the output but i does show the same again and again.

8,0    2      143    10.099853386  1593  I  WS 1918047520 + 8 [kvm]
  8,0    2      144    10.099868014   198  D  WS 1918047520 + 8 [kworker/2:1H]
  8,0    2      145    10.100108188  1593  Q  WS 5418165048 + 8 [kvm]
  8,0    2      146    10.100109169  1593  G  WS 5418165048 + 8 [kvm]
  8,0    2      147    10.100109430  1593  I  WS 5418165048 + 8 [kvm]
  8,0    2      148    10.100143975   198  D  WS 5418165048 + 8 [kworker/2:1H]
  8,0    2      149    10.100622299  1593  Q  WS 5475189208 + 8 [kvm]
  8,0    2      150    10.100623201  1593  G  WS 5475189208 + 8 [kvm]
  8,0    2      151    10.100623461  1593  I  WS 5475189208 + 8 [kvm]
  8,0    2      152    10.100637167   198  D  WS 5475189208 + 8 [kworker/2:1H]

Also it does not make any difference if the VM which is using the drives is online or not (TrueNAS Core VM)

Those drives are able to spin down and sleep. If i send the command

hdparm -Y /dev/sda

they go to sleep. Only weird thing is that i cannot get any state when i try the command

hdparm -C /dev/sda

i just get "drive state is: unknown. Anyways. I used those exact HDD's in my Synology NAS where they went to sleep and did wake up when they got some input from Proxmox - Jellyfin Container.

Does someone has any idea? Please help, i am out of ideas and pretty new to the Proxmox environment. I am happy to share some additional information if needed

 

[gfngfn256]

Physical HDD's (besides some others) attached as Data Storage

If those are setup within the PVE storage system, the drives aren't going to sleep as PVE polls these drives ever so often for data metrics.

hdparm -C /dev/sda

Have you tried this (as per docs) instead:
hdparm -I /dev/sda or maybe this smartctl -i -n standby /dev/sda

 

[meyergru]

Even if Proxmox would not scan those drives periodically, ZFS on TrueNAS will do it anyway.