RP5470 Hung?

[James Raffeld]

just an update to those who may be following this thread. I am using this as a learning experience and will call HP if needed. So far I have now been able to reset the system and interupt the autoboot and do a search for potential boot drives and the 18gb drive at 0/0/1/1.0 does not show up this drive is one of 2 drives in vg00 and i a, able to boot from the install cd with no errors.

ran ioscan and neither drive on the 0/0/1/1 bus showed up. Reseated both drives and now both show in ioscan.

Rebooted system and did a search for boot devices an the drive at 0/0/1/1.0 now shows up as a potential boot device.

booted system from promary boot device and system booted without errors.

OLDsyslog.log had following errors:



Nov 18 11:44:22 larry EMS [2182]: ------ EMS Event Notification ------ Value:
"MAJORWARNING (3)" for Resource: "/storage/events/disks/default/0_0_1_1.0.0"
(Threshold: >= " 3") Execute the following command to obtain event details:
/opt/resmon/bin/resdata -R 142999554 -r /storage/events/disks/default/0_0_1_1
.0.0 -n 142999553 -a
Nov 18 11:44:23 larry EMS [2182]: ------ EMS Event Notification ------ Value:
"MAJORWARNING (3)" for Resource: "/storage/events/disks/default/0_0_1_1.2.0"
(Threshold: >= " 3") Execute the following command to obtain event details:
/opt/resmon/bin/resdata -R 142999557 -r /storage/events/disks/default/0_0_1_1
.2.0 -n 142999554 -a
Nov 18 15:45:53 larry : su : + 1 root-oracle
Nov 18 15:41:54 larry vmunix: SCSI: Target detected parity error -- lbolt: 14877
91, dev: 1f010000
Nov 18 18:32:50 larry vmunix: SCSI: Unexpected Disconnect -- lbolt: 2513351, dev
: 1f010000, io_id: 1015400
Nov 18 18:33:17 larry vmunix: SCSI: Resetting SCSI -- lbolt: 2515609, bus: 1
Nov 18 18:33:17 larry vmunix: SCSI: Reset detected -- lbolt: 2515609, bus: 1
Nov 18 18:33:17 larry vmunix: LVM: Recovered Path (device 0x1f010000) to PV 1 in
VG 0.
So i would have to conclude that the disk at 0/0/1/1.0 is bad.

Any thing else to add or differing opinions?

Thanks so much for the help and i must say using these forums for is such a great help!

[Bryan D. Quinn]

Hello James,

Looks like you nailed it. I would definately still consider having HP swap out that drive.

-Bryan

[Bernhard Mueller]

James,

good you got it up again, however, you got EMS warnings at about the same time from *two* disks.

Unlikley two drives fail at the same moment, maybe both were not seated properly (did you by chance experience any earthquakes lately ;-)

Still to be on the safe side regarding the disk drives, do the following.

run diskinfo /dev/rdsk/cxtYd0
it should return immediately, longer pausing indicates a problem even if ioscan shows CLAIMED.

even if diskinfo is ok, run a full
dd if=/dev/rdsk/cXtYd0 of=/dev/null
this should take a while and exit saying
# records in
# records out

If you get an I/O error from dd replace the drive.

Regards,
Bernhard

[Bernhard Mueller]

oops

to speed it up a bit use
dd if=/dev/rdsk/cXtYd0 of=/dev/null bs=1024k

Though as mentioned, I would assume the drives to be good and rather suspect a backplane problem.

Regards,
Bernhard

[James Raffeld]

Thanks Bernhard, had another drive and put it same position so that it would be at 0/0/1/1.0 and issued a seach for boot devices and new drive did not show up as a potential boot device. So I like believe it is not a disk drive. Calling HP! Thanks!

[Michael Tully]

While your on the job, this would be a very oppotune time to update your GSP and CPU firmware. Having the latest patch bundle loaded should have been a pre-requisite to doing these as well.