EMS reports failed disk on Clariion every 24 hours

[John Jeffrey]

I have an existing SAN with a DMX, and we're migrating to a Clariion. I have my first HP-UX test host connected to the Clariion. Everything seems to be working fine, but EMS reports a failed disk to the path of the non-owning SP every day at the same time. I have a call open with EMC, but they're clueless.

I'm running HP-UX 11.0.

Event data from monitor:

Event Time..........: Wed Jul 25 17:27:53 2007
Severity............: SERIOUS
Monitor.............: disk_em
Event #.............: 100472
System..............: python.ltd.amphenol-tcs.com

Summary:
Disk at hardware path 1/10/0/0.97.50.19.0.0.5 : Device connectivity or
hardware failure

I have Powerpath 4.2.0, which is a bit old, but shouldn't really matter.

Currently, I can't get the naviagent to run. I guess it broke during troubleshooting or something. (EMC is clueless on this as well.) However, it WAS running when the problem began, so I don't think the fact that it currently won't run is the cause of the issue.

The settings on the Clariion initiator ports are correct, per EMC. (HP No Auto Trespass, Array Compath Enabled, Failover Mode 1)

Trespassing the lun will cause the alert to report on new non-owning SP. Again, every day at the same time (about 16:30). Not at the time of the trespass. I actually added a batch of more luns, and the event for them all fire about an hour later (about 17:30).

I don't have anything cron'd that would be doing any low level scanning. Just the standard HP-UX tools. EMS, STM, and whatever else.

I cannot trigger the event manually. Ioscans, or forcefully trespassing the lun from the host do not trigger it.

Thanks in advance.

[Duncan Edmonstone]

Well if I recall correctly, EMS is supposed to completely ignore all non-HP disks (on the grounds it doesn't know how to interrogate them correctly anyway).

I would make sure you are on the latest and final version of Onlinediags (STM) for 11.00, which I think was B.11.00.27.18. You can check this by running

cstm

from the command line.

However if you're not on the latest version I really don't know where you will get it from if you can't lay your hands on the CDs, as 11.00 is now obsoleted and completely out of support (you *do* know that right?) and they are no longer downloadable from any HP sites (at least that I can find).

I believe this release was on the March 2004 HP-UX 11.00 Support Plus CDs if you can lay your hands on that.

Of course if you're already on that version then there's not a lot more HP are going to do for you what with 11.00 being out of support. I seem to recall there is a way of manually excluding LUNs from disk_em polling them - let me know if you need to go down that route and I'll try and dig out some old notes.

HTH

Duncan

[skt_skt]

Here is the version , i have for STM

-- Information --
Support Tools Manager
Version A.35.10

"Well if I recall correctly, EMS is supposed to completely ignore all non-HP disks (on the grounds it doesn't know how to interrogate them correctly anyway"

i dont agree to this . I have the clar disks whihc are getting monitored thorugh EMS too.

[Duncan Edmonstone · Accepted solution]

Oops - I was giving you the STM version information as reported in swlist for onlinediag. The version reported by STM when run will be A.44.0 if you have the most recent version for 11.00.

Also John, looks like I was wrong about the latest version of diagnostics for 11.00 *is* still available here:

http://h20293.www2.hp.com/portal/swdepot/displayProductInfo.do?productNumber=B6191AAE

There's also a patch for that version of STM - PHSS_34834.

HTH

Duncan

[Andrew Merritt_2]

Hi John,
Duncan has pretty much covered the answers. If you are running A.44.00 with PHSS_34834 and you are still seeing the problem, please contact HP support.

Santhosh, you need to upgrade. A.35.00 was the December 2002 release, and had the problem Duncan mentions, that it was monitoring non-HP disk devices. A.44.00 is the only supported release for 11.00.

http://www.docs.hp.com/en/diag/stm/stm_upd.htm#table shows the various releases.

Andrew

[Mark Landin]

"A.44.00 is the only supported release for 11.00."

You mean *was* the only supported release, right?

[John Jeffrey]

Thanks guys! Upgrading to v44 of Support Tools fixed it.

And yes, I know 11.0 is desupported. Unfortunately, it's not my decision. The server is running critical apps that are estimated to cost ~$1mil to upgrade.

[Andrew Merritt_2]

John,
Good to know that fixed your problem, and thanks for letting us know.


>> "A.44.00 is the only supported release for 11.00."

> You mean *was* the only supported release, right?

There are many levels of support. See http://www.hp.com/softwarereleases/releases-media2/history/slide2.html - even though HP-UX 11.00 itself is out of factory support, "limited support may be offered to an additional duration"; on top of which some customers pay for extended support. For customers with a support contract, calls on A.44.00 OnlineDiags on 11.00 will still be taken, and investigated, though the chances of patches being produced to fix any newly discovered defects are vanishingly small. Problems reported on earlier versions are likely to be met with a request to upgrade to the latest version.

This is under review, and at some point in the next few months support for this version of OnlineDiags will likely cease too.

Andrew