[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

AS1000A hardware issues...



This isn't so much a Debian-alpha question as it is a general alpha
question, but if I can get these issues resolved the box will have Debian
on it again =)

The hardware is an AlphaServer 1000A 4/266 with a onboard QLogic ISP
1020, DE500 NIC, Trio64v+ video (onboard video disabled) and Mylex
DAC960PL RAID. I have SRM 5.7-83, which I believe was the last one
released for this model.

Basically the onboard QLogic is the problem, I have a Seagate ST34371N
(id 0) and the DEC RRD45 CDROM (id 4) plugged into it and they're both
giving me problems. The cdrom problems could be because I'm using CDRs,
but IIRC they worked fine before. I'm getting sparse errors, sometimes
SRM will see the seagate right away sometimes it takes a minute,
sometimes it boots up all the way sometimes it freezes in the middle of
loading a kernel via aboot, if it boots all the way it usually oopses
after a bit of work.

After a cold power on I will sometimes get this in the SRM error log
(seen via `cat el` if I'm referring to it incorrectly):
I snipped the middle and right column because I believe they're just
other representations of the same data, the 00143 appears to always be
the same but the rest sometimes changes although they're always in order.

I/O failed on pka0.0.0.2000.0 
00143b0c
00143b1c
00143b2c
00143b3c
00143c0c
00143c1c
00143c2c
00143c3c

Other times I'll get something similar to : "mailbox I/O failed for
pka0.4.0.2000.0" with a state number after it, 8006 comes to mind but I
don't have that error infront of me right now to verify. I don't believe
the drive is messed up because I hooked it up to one of my x86 boxes
1542's and ran badblocks on ~25% of the device (couldn't mount it because
I didn't have support for OSF disklabels in the kernel)  and it had no
problems. I don't believe it's the cable because I tried another cable,
the one from the 1542 actually, and I got the same erratic symptoms. Is
there someway I can verify the QLogic is working ok? if I run disks
(which appears to be a script that runs exer on the scsi disk) it
eventually fails with a hard error but that doesn't tell me what's messed
up.

I just got it to oops and it blamed klogd, the code with an * is "stl
zero,0(zero)" I can post the whole thing if need be, but it's alot to
type =)

I'm going to try to snag a spare disk from work tomorrow, because it's
possible there is shock damage since I had to physically transport the
box recently, and I'll send again if that fixes my problems but I have my
doubts

Anyway I'm going to get some sleep, thank for any help you can provide.
-- 
Help protect your rights on-line.
Join the Electronic Frontiers Foundation today: http://www.eff.org/join
-----------------------------------------------------------------------
the death of one is a tragedy, the death of millions is just a statistic




Reply to: