# dmesg | grep ad4
ad4: 953869MB <WDC WD10EARS-00Y5B1 80.00A80> at ata2-master UDMA100
GEOM: ad4: partition 1 does not start on a track boundary.
GEOM: ad4: partition 1 does not end on a track boundary.
GEOM: ad4s1: geometry does not match label (32h,32s != 16h,63s).
Trying to mount root from ufs:/dev/ad4s1a
# smartctl -a /dev/ad4 | grep Load
193 Load_Cycle_Count 0x0032 188 188 000 Old_age Always -
38319
Is there any way to turn off or increase the 8-second timeout? Or do I have to
bin the disk and use something better?
I don't want to boot into DOS, windows or any of that stuff. I also tried
ataidle to no effect.
Cheers
Michiel
All I can suggest is a small daemon that just forces the disk to do
physical I/O every couple of seconds.
--
Peter Jeremy
I guess 'hdparm -B' is equivalent to 'ataidle -P'? In this case I'm out of luck:
# ataidle -P 254 /dev/ad4
ataidle: the device does not support advanced power management
Cheers
Michiel
>Hi. I have one of those new green WD HDs with Intellipark, that is, they
>like to park their heads every 8 seconds. As a result the load cycle count
>grows at an alarming rate. (If I understand correctly this number should not
>exceed about 300k or so.)
>
>Is there any way to turn off or increase the 8-second timeout? Or do I have
>to bin the disk and use something better?
I have some older GP drives which do not seem to ever park their heads. The
newer GP drives I have behave as you described.
I have had success either turning off the head parking or greatly reducing it
by doing 'hdparm -S 0 -B 254 <device>'. Either '-B 254' or -B 255' depending
on the drive. I use these drives in a home file server, though, so I/O
to/from the drives is either non-existant most of the time (heads stay
parked) or there is continuous I/O (rdiff backups, etc.; heads stay
unparked).
-akm