I have a RedHat 6.0 system that I've been running for some time now. Recently, I installed another 4GB drive (Western Digital) in it (it started out with a Seagate 4GB drive). Since I installed the new drive, the system has become sluggish, and every minute or so it ehcos the following to the first virtual console:

hdb: irq timeout: status=0xd0 { Busy }
ide0: reset: success

Every time this happens, I can hear the drive park and stop after the first line of the message, then spin back up before the second message. Both drives are accessable, but the sluggishness is a problem, and I'm a bit worried about wear on the drive from repeated stop/park/spin up operations.

I would like to know if there is a nice, canned silver bullet for this, or at least have some idea what the error means.

Im guessing this is the case where you have a newer mobo and an older HD, and have DMA enabled. Executing "hdparm -d /dev/(drive)" would tell you. Executing "hdparm -d 0 /dev/(drive)" turns off DMA for that drive.

HTH

Checked both drives with hdparm. They both have exactly the same settings, with DMA off. Output of hdparm for the offending drive follows:

multcount = 0 (off)
I/O support = 0 (default 16-bit)
unmaskirq = 0 (off)
using_dma = 0 (off)
keepsettings = 0 (off)
nowerr = 0 (off)
readonly = 0 (off)
readahead = 8 (on)
geometry = 523/255/63, sectors = 8404830, start = 0

Any other ideas?

have same problem with old WD caviar 31000..

do you have already a solution ??

Victory! Here's what worked for me (keep in mind that the culprit drive is a Seagate ST34321A 4GB - my /dev/hdb)

hdparm -S 250 /dev/hdb

This sets the spin down time to 5 hours. See man hdparm for more info - the -S flag is a bit odd in the way it works.