SCSI problem help needed identifying disk

[Tony Horton]

OK Maybe third time lucky the last one was the same as the earlier one.

Would a pvdisplay of the disks be usefull?

[Michael Steele_2]

Not a lot going on except this:

18:40:19 30 10 20 40

But nothing on c6 but:

18:36:13 Average c6t12d0 0.04 0.50 0 1 3.09 8.53

Use 'pvdisplay' to see what logical volumes are on c6 disks and cross ref. to 'bdf' output.

[Tony Horton]

Hi Michael,

I don't think the problem is load related. I have an L1500 which i have been experimenting with before swapping it with the box above that has the problem. The other day it came up with the same messages in the log. Funny thing is, it was on a disk which I have added to a volume group that doesn't have any lvols defined on it, and the box was completely idle, ie it has the os installed on it and nothing running at all. The disk was simmilar although not identical to one of the ones giving a problem in the L1000. Its a maxtor atlas 10K IV 73GB (the L1000 has Atlas 10K III 73GB).

HP have released firmware updates for pretty much the problem I'm seeing (for their hp branded Atlas 10K III drives), so I guess what I really need to do is talk to maxtor and see whether they are willing to do anything about it.

Also I'm no longer certain that the pauses are related to the disk messages. apparently the pauses have been much more frequent than the appearance of messages. The machine has a bit of a memory problem, so it may be unrelated.

I did have a linux box with one of these atlas 10K III drives which was behaving similarly (ie locking up for 20 to 30 seconds), eventually the disk developed a whole pile of bad sectors and when I replaced it the system stopped locking up. I'm suspicious this is the same scenario.

Regards,

Tony.