I have a WD4000F9YZ harddisk connected to the 2nd sata port of a intel atom d510mo mainboard.

The drive will go to standby with the -y flag, but when setting timeout with -S the drive will not go to standby after the timeout period.

I also have an SSD connected to the first sata port of the mainboard, and that actually does go to standby after the timeout period (not that this saves any power though).

I contacted Western Digital about this problem, and they told me that not any harddisks have the option to set a timeout value. The timeout is done by the OS or the RAID controller which will send a standby command to the drive after the timeout occurs.

I found a script that runs with cron every 5 minutes and when there is no disk activity it sends the standby command with hdparm -S but this is a work around.

Where is the timout handling done? Is it in kernel or is in in the ICH7 chipset on the mainboard?

Why does the standby timeout not work on this drive?

Kernel version: 4.8.0-39-generic

Ubuntu version: Ubuntu 16.10 (updated to latest packages on 1-3-2017)

Hdparm version: v9.48

The timeout is done internal to the drive, and "hdparm -S" just sets an internal parameter. As the WD rep. told you, not all drives (especially not all WD drives) support that feature.

Thanx for the answer. Now it's clear to me that it is a parameter set in the HDD and not a timer in the kernel or SATA controller.

After some googleing I found out that WD's indeed can have lack of support for standby timeout :-(

http://withblue.ink/2016/07/15/what-...-home-nas.html

For drives that don't support the timeout parameter a daemon is written on: http://hd-idle.sourceforge.net/

I also found this script that can run as a cronjob:

http://superuser.com/questions/65188...x-raspberry-pi