I'm currently configuring a new home server, and I am trying to get the disks to spin down after one hour. But they do not spin down.


I have my entire system on an ssd: /dev/sda. I have three hdds in the system: /dev/sdb, /dev/sdc and /dev/sdd. They are not mounted, although they do contain (active) LVM volumes.

What does work

I can put the disk into standby using :

hdparm -y /dev/sdb

The disk stays spun down for days (if I leave the system alone), indicating (I think) that the system really isn't generating any activity

Using relatively small values for the timeout like :

hdparm -S 180 /dev/sdb
hdparm -S 240 /dev/sdc
hdparm -S 241 /dev/sdd

also works. Each of the disks spins down after the specified amount of time (15 minutes, 20 minutes and 30 minutes, respectively)

What doesn't work

I try to set the timeout value to one hour, like so :

hdparm -S 242 /dev/sdb

The command appears to succeed, but the disk doesn't go into standby at all (i.e. after a day, the disk is still in active/idle mode, as witnessed by my power meter, as well as

hdparm -C /dev/sdb

System details

This is Debian Squeeze, running on an Atom-powered Asus motherboard (AT5NM10T-I). The disks are:

[1.948217] scsi 0:0:0:0: Direct-Access ATA OCZ VERTEX PLUS  3.50 PQ: 0 ANSI: 5
[1.953713] scsi 1:0:0:0: Direct-Access ATA ST2000DL003-9VT1 CC3C PQ: 0 ANSI: 5
[1.954302] scsi 4:0:0:0: Direct-Access ATA ST31500341AS     SD1B PQ: 0 ANSI: 5
[1.954873] scsi 5:0:0:0: Direct-Access ATA ST31500341AS     SD1B PQ: 0 ANSI: 5

migrated from serverfault.com Jan 10 '12 at 17:25

This question came from our site for system and network administrators.

  • @syneticon-dj: Be careful with this suggestion, some people tend to create new questions on the suggested site(s) instead of waiting for their question to get migrated by votes. Kees-Jan: Please don't do that :) – Sven Jan 10 '12 at 17:13
  • Just want to add, shouldn't it be hdparm -S 720 /dev/sdb for one hour standby? – Cold T Jan 10 '12 at 17:13
  • @ColdT: Nope, it shouldn't be. According to the hdparm manual "The encoding of the timeout value is somewhat peculiar". See the manual for the full explanation (linux.die.net/man/8/hdparm) – Kees-Jan Jan 10 '12 at 17:45

In its default configuration, smartd checks disk status once every 30 minutes, except when the disk is in standby.

Apparently, that generates enough disk-activity for disks to never actually go to sleep (for timeouts larger than 30 minutes), but doesn't wake-up any sleeping disks.


You should try if it works when all partitions on that disc ar unmounted. I also fiddled with this a while ago and i had some troubles with regular reads/writes holding the disc in spinning state.

On the other hand see this:

Power cycling control
Shutting down and rebooting a computer or resuming it from hibernation cycles the power to the drives in the computer. The spin-up operation performed by a drive after a power cycle is believed to place more stress on the drive than running the drive continuously for a long period of time.

Based on professional experience of system administrators, it is believed that there is a direct relationship between the number of power cycles of a computer and the probability of failure of its drives. In other words, a computer with a high uptime may have a lower probability of drive failure than one that has its power cycled routinely.


  • Thanks for your answer. In my test setup, the disks are not mounted (as mentioned in the setup section). I am aware that people believe that powercycling the disk causes stress, but we're talking here about a disk that is rarely used (less than once a month), and I do have a power bill to pay. – Kees-Jan Jan 12 '12 at 22:26

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.