HDD Spindown not working?

maexle

New Member
Mar 10, 2026
1
0
1
Hello to everyone,

I am running Proxmox in my Homelab and have tried to configure a spindown for two HDD, but it is not working the way I want it to.

Both HDD's are configured in the /etc/hdparm.conf file with a spindown timer set to 20 minutes. Also one of the drives supports APM which is set to apm = 127.
The small HDD with 500GB from WD (no APM-Support) is configured as a zfs-volume through the WebUI to provide storage for multiple Container's/VM's. My Problem is that the this drive is not spinning down. There is no running Container's/VM's which can access the drive.

I have looked up on the Internet how to solve the problem without solving the problem.
One possible solution I tried was to edit the /etc/lvm/lvm.conf file.
At the end of the file I added the both drives to global_filter.
Code:
devices {
     # added by pve-manager to avoid scanning ZFS zvols and Ceph rbds
     global_filter=["r|/dev/zd.*|","r|/dev/rbd.*|","r|/dev/sdb.*|","r|/dev/sdc.*|"]
}
There is also global_filter part above in the file but not active.
I tried different things with the file, rebooted the machine and ran vgscan to reload the settings without any effect.

Did I do something wrong while editing the file or is there something else that cloud stop the thing from working correctly?

kind regards
maexle

PS. My english is not perfect
 
Similar problem here.
You can try hd-idle and verify if the disks are sleeping but are awaking every 10 minutes like me

Maybe you can also put 5 mins instead of 20 mins sleep timeout and check if they sleep but wake up every 10 mins
 
Last edited:
The small HDD with 500GB from WD (no APM-Support) is configured as a zfs-volume through the WebUI to provide storage for multiple Container's/VM's. My Problem is that the this drive is not spinning down.
Proxmox checks the used/free space of each (enabled) storage all the time, keeping it spinning. There are existing threads about this but I don't know if (or how) you could change that.
 
  • Like
Reactions: news and Johannes S
Relevant threads:
https://github.com/openzfs/zfs/issues/18082
https://github.com/openzfs/zfs/pull/18138

There is a bug at openzfs 2.4.0 that make disks wake up every 10 minutes (or do not let them sleep if the sleep timeout is less than 10 minutes)
This should be fixed at openzfs version 2.4.1.
Temporary fix (according to the above thread) :
echo 31557600 > /sys/module/zfs/parameters/spa_flush_txg_time
(this does not survive reboots)

or put :
options zfs spa_flush_txg_time=31557600

in /etc/modprobe.d/zfs.conf
The zfs.conf solution survives reboots

Remember to delete it after openzfs 4.1 update