On Tue, Aug 08, 2000 at 08:48:25PM +0200, Marc SCHAEFER wrote:
> On Tue, 8 Aug 2000, Ricky Beam wrote:
>
> > Technically, one could argue the kernel should issue a STOP START UNIT command
> > for you. A command was sent to the drive that needs it to be spinning (and
>
> Yes, infact if you apply the scsi-idle patch, I think it does. However,
> last time I used that was in 2.0.27 ...
>
> > the kernel is assuming it is spinning??? is there any state tracking?) Of
> > course, one could argue the kernel should only do what it's told...
>
> Also, on bootup (on sd_mod insertion), disks are spun up.
This sounds like a good time for me to mention another issue I'm seeing...
Castlewood makes a device called an ORB drive. Upon a long period of
inactivity, the drive spins down. This wouldn't be a problem, except that
when the next media-access command occurs, it fails with a CHECK_CONDITION
and returns sense data that indicates a power-on reset. This causes all
sorts of SCSI I/O errors.
It appears that this is another condition where the kernel is assuming that
the drive is still spinning. I'll admit, this assumption _should_ be
valid, but what can I say....
Does anyone have any ideas on how this could be handled? I see two obvious
solutions: send keep-alive read commands, or send an explicit start-stop
when the activity stops (or restarts).
Matt
--
Matthew Dharm Home: [EMAIL PROTECTED]
Senior Engineer, QCP Inc. Work: [EMAIL PROTECTED]
G: Baaap booop BAHHHP.
Mir: 9600 Baud?
Mik: No, no! 9600 goes baap booop, not booop bahhhp!
-- Greg, Miranda and Mike
User Friendly, 12/31/1998
-
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to [EMAIL PROTECTED]
