Unexpected SCSI busfree messages then server hangs 1

[Tarquin]

We've been having this issue intermittently since we first installed Linux and it is still more than annoying:

Oct 9 03:17:57 mel kernel: (scsi0:0:5:-1) Unexpected busfree, LASTPHASE = 0xe0, SEQADDR = 0xcc

Oct 9 03:17:57 mel kernel: (scsi0:0:15:7) Invalid SCB during SEQINT 0x71, SCB_TAG 255.

Oct 9 03:17:57 mel kernel: (scsi0:0:6:0) No active SCB for reconnecting target - Issuing BUS DEVICE RESET.

Oct 9 03:17:57 mel kernel: (scsi0:0:6:0) SAVED_TCL=0x60, ARG_1=0x1, SEQADDR=0x102

Oct 9 03:17:58 mel kernel: (scsi0:0:6:0) Synchronous at 5.0 Mbyte/sec, offset 15.

The messages will then repeat, nothing else will process, until the server is hard booted. (In this instance the server had been running happily for 28 days.)

The server (Dell 2400) at this time is just starting the backup job using a Dell badged Seagate DDS4 drive attached via an Adaptec AIC7880 controller.

We've had this problem under RH 6.2, 7.0 and 7.1 .

Looking back through web mail archives this issue has cropped up for several people many times for at least 2 years! (with various kernel versions)

What is the cause, what is the resolution? I suspect it's driver but Adaptec controllers are not rare.

Thanks in advance.

 

[ifincham]

Hi,

If you look, there are similar problems with people on freebsd & other bsd variants so its not limited to linux. However, I believe the drivers come from essentially the same source so maybe there are issues there. You could try a newer version from direct from -->

http://people.freebsd.org/~gibbs/linux/

.

Regards

 

[Tarquin]

Thanks, we'll move away from RH standard.