SCSI problems with Qlogic ISP
I've been seeing a lot of scsi errors recently on some of the drives
attached to an alpha we have. All of these appear to be sense/parity
errors of the form:
scsi0 channel 0 : resetting for second half of retries.
SCSI bus is being reset for host 0 channel 0.
SCSI disk error : host 0 channel 0 id 2 lun 0 return code = 18000002
[valid=0] Info fld=0x0, Current sd08:23: sense key Aborted Command
Additional sense indicates Scsi parity error
scsidisk I/O error: dev 08:23, sector 8913048
We're running 4 drives + one tape drive off a QLogic 1020 ISP adapter
(via a wide->narrow adapter). The total cable length exceeds 1.5m
which I've seen elsewhere on the web as being the max length for this
set up. The whole bus is terminated with an active terminator and the
QLogic adapter is set to auto-terminate. We're running Debian 2.2,
kernel 2.2.14 with the qlogicisp driver.
Are these sort of errors indicative of cabling/termination problems or
should I be looking elsewhere? If we swap around/replace the cables
the drive id causing the error changes so I cant believe its a drive
error. What worries me is the whole machine locks up solid after these
errors and I dont know whether these errors should cause that...
Marcus Williams - http://www.onq2.com
Quintic Ltd, 39 Newnham Rd, Cambridge, CB3 9EY