2.4.24 AIC7xxx kernel messages

From: Paul Armor
Date: Fri Mar 05 2004 - 08:02:23 EST


Hi all,

Please include me in responses, as I'm not currently subscribed.

Googling and searching various archives, I can't find more than passing
references to my problem, so I'll ask here. Has anyone else seen anything
like this, or can anyone offer suggestions as to what may be wrong? I've
got an Adaptec 29160, off of which I've got 4 Promise UltraTrak SX8000's,
across which I'm striping software RAID. Sometimes, when under heavy
load, the kernel will log lot's of messages, starting with:

Mar 5 03:47:15 storage2 kernel: scsi0:0:0:0: Attempting to queue an ABORT
message
Mar 5 03:47:15 storage2 kernel: CDB: 0x28 0x0 0x27 0xaa 0xe5 0x57 0x0 0x0
0x10 0x0
Mar 5 03:47:15 storage2 kernel: scsi0: At time of recovery, card was not
paused
Mar 5 03:47:15 storage2 kernel: >>>>>>>>>>>>>>>>>> Dump Card State Begins
<<<<<<<<<<<<<<<<<
Mar 5 03:47:15 storage2 kernel: scsi0: Dumping Card State in Data-in
phase, at SEQADDR 0x54
Mar 5 03:47:15 storage2 kernel: Card was paused


...after which it will list SCB stats and info.
This last occurance started after the machine had been copying
data from one striped partition to another (it was ~250GB/1.1TB complete),
and the log messages stopped ~11 minutes later; and the cp has continued.
It seems as though the card/kernel/external arrays will work well together
for many hours, get into this state for a few minutes, then work for many
more hours, etc... Any comments or suggestions would be GREATLY
appreciated.

Here is the config:

> motherboard - Asus A7M266
>
> kernel - linux-2.4.24
>
> controller - Adaptec AIC7xxx driver version: 6.2.36
> Adaptec 29160 Ultra160 SCSI adapter
> aic7892: Ultra160 Wide Channel A, SCSI Id=7,
32/253 SCBs
>
> ext. RAID arrays - 4 x Promise UltraTrak SX8000
> F/W version 1.1.0.30 (2/4/2003)
>
> Each box has three partitions
>
> We're running software raid 0 striping across the 4 boxes such that
> md0=sd[abcd]1, md1=sd[abcd]2, md2=sd[abcd]3.

-
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/