Re: USB attached SATA disk idle spindown

[Stefan Richter]

(adding Cc: linux-usb and Johannes Niess)

On 6/24/2008 6:59 AM, Jonathan Steinert wrote at LKML:
> Dear list,
> 
> I am looking for what the current state of USB SATA timed spindown of-
> disks is ata this time. I've been reading documents and mailing lists-
> for every possible search I can dream up on the subject. Suggested-
> workable solutions include (but are probably not limited to):
> 
> $ hdparm -S
>     This seems to be the ATA way of telling a disk to spin down, and it
>     works in some situations, but doesn't appear to work on USB attached
>     disks under 2.6.25
> 
> $ sdparm
>     This is suggested in a small number of locations. Most of which have
>     examples on how to disable STANDBY on a handful of devices. I am
>     unable to find any evidence that it can actually be used to enable
>     power saving.

I think either of these depend on specific support by the USB to SATA
bridge firmware.

I hear some newer bridges and their firmwares implement ATA
pass-through.  Would that help?

> $ scsi_spin
>     Requires the disk to be unmounted, and I can't find anything saying
>     that this is how the world is supposed to be.-

Johannes recently informed me that he wrote a script which is used like
a background daemon to stop idle USB disks.  The script uses
  - "cat /sys/block/${device}/stat" to watch disk activity,
  - "sdparm -q -C stop /dev/${device}" to stop the motor after
    longer inactivity.

> I'm hoping that perhaps I'm missing something from this list.
> 
> I don't really know whether this is a topic for the scsi people, or the-
> SATA people, or the USB people. So I'm asking here instead to try and
> get my first hints from the parent group.
> 
> Thanks in advance,
> 
> hachi

IMO it is firstly an USB topic because it depends on the features and
behaviour of the bridge and its firmware.

(PS:  Don't have too high expectations of $15 devices.)
-- 
Stefan Richter
-=====-==--- -==- ==---
http://arcgraph.de/sr/
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/