From: Michael Bartosh (mbartosh@4am-media.com)
Date: Tue Dec 17 2002 - 22:22:55 EST
Not really a summary, since I'm hoping for more, but:
In brief: D1000 array doesn't work, and I'm seeing strange scsi bus
resets. (Solaris 8)- 3 disks weren't lighting up.
Suggestions
-is the card compatible? (yes, it's X6541A)
-can you see the disks with format? (no)
-try: ok probe-scsi-all (results in hang)
-is the terminator HVD (Ah.. no! it's not!)
Well, setting the bank of switches on the back to this position got all
the disks to spin up:
1 up
2 down
3 down
4 up
5 down
...but this isn't the default position, is it?
Clue?
What address is the box supposed to be set to (the round dial?)
Unfortunately, after ordering an HVD terminator, I still only see one
disk with probe-scsi-all.
Errors in /var/adm/messages look like:
WARNING: /pci@1f.0pci@5/scsi@2 (glm0)
Connected command timeout forTarget 0,0
glm0: fault detected in device still available
got scsi bus reset
glm0: fault detected in device still available
sd1 at glm0: target 0 lun 0
WARNING:
corrupt label - wrong magic number
sd2 isn't mentioned but I also have corrupt label- wrong magic number
warnings for sd3, sd4, sd5, sd6, sd8 (where's 7?), sd9, sd10, sd111,
sd12 and sd13.
After running devfsadm (even though prob-scsi-all puked) format sees 11
of the 12 disks in the array.
Thanks:
Joon Martin Hansen
Ric Anderson
Justin Stringfellow
Peter Stokes
Darren Dunham
Matthew Brown
Tony Urban
On Sunday, December 15, 2002, at 01:58 AM, Michael Bartosh wrote:
> I recently had a D1000 given to me.
>
> Mostly, I use solaris for interoperability testing (iPlanet Dir Srv
> and MIT KDC) since that's what one of my largest customers uses. But
> I'll take a large array any day ;-)
>
> So I don't have much of a clue as to how to get it up and running, and
> google is failing me. The OEM'd card seems to be recognized fine (Rev
> 5 53c875)- but then we have lots of glm0 errors-
>
> kern.info NOTICE: glm0: fault detected in device; service still
> available
> ...got scsi bus reset
> ...connected command timeout for Target 4.0
> ... for Target 6.0 (through 14.0)
>
> (assuming glm0 = scsi bus A, to which the array is connected)
>
> The array appears to be configured according to the PDF I downloaded-
> ie 2 busses linked together, and one terminated. The lights
> corresponding to three of the disks, however, aren't on (there are,
> however, disks present)
>
> Ultimately I'd like to move the array to another platform, since my
> blade is fairly important for testing, and I need to reconfigure it on
> a regular basis. I figured getting it up on solaris would be a good
> first step though.
>
> Any clues or pointers?
>
> Thanks-
_______________________________________________
sunmanagers mailing list
sunmanagers@sunmanagers.org
http://www.sunmanagers.org/mailman/listinfo/sunmanagers
This archive was generated by hypermail 2.1.7 : Wed Apr 09 2008 - 23:25:29 EDT