after replacing emc disks, powermt check cmd HANGS

[Steve Post]

CX300, Powerpath, hpux11.23

We have 2 luns that belong to a non-mirrored raid group. The raid group only holds onto one really big backup disk. Anyway, we had to replace a bad physical disk. This meant I had to unbind the 2 luns and rebind them, empty, to this raid group.

I would rebuild the volume group that was destroyed. But I can't get there yet. I have a small problem. The disks have the wrong device id's.

The command powermt check is hung.
I see the old luns are still seen by power path. But their Device IDs are not the same as the new LUNs device IDs.

I want to know.....

how I can tell power path to assign new device id's to these two newly resurrected luns?

Would a reboot fix it?


The powermt check command sees the busted lun and asks if I want to remove it. I said yes. Then it HUNG.

[Rita C Workman]

cx series is Navisphere I believe.

You might get more assistance trying the EMC Forums site:

http://powerlink.emc.com

Support/Support Forums

If it were Solutions Enabler, I could offer some help. But it has been too many years since I last touched Navi..

Sorry & hope this helps,
Rita

[Steve Post]

been there. done that.
I submitted a support ticket over an hour ago with EMC. I haven't gotten an answer. HP forums usually gives me a faster and better response than any other support I have seen.

I'm not saying support is BAD. I'm saying hp forums are GOOD. So I normally pose the question here.

I can guess.... I think I have to reboot. Other forum entries talk about running syb??? command. But I do not even have it.

Guess 1: run powermt check command on a different session.
powermt display
powermt check
ioscan -fnC disk
insf -e
powermt config
powermt display
powermt save


Guess 2: reboot

powermt display
powermt check
ioscan -fnC disk
insf -e
powermt config
powermt display
powermt save


I know that currently the 2 LUNs have the wrong, old "logical device ID". This number looks a lot like World Wide Name.

[Steve Post]

I got my answer.

mv /etc/powermt.config /etc/powermt.configsave
powermt check force
ioscan -fnC disk
ioscan -fC disk
(oh...one of those)

powermt config
powermt display
powermt save


IF the powermt check command locks again during this? Reboot and try the same commands again.

And yes. That is what happened. The powermt_check_force command locked because it was waiting for my older powermt_check command to complete. The reboot fixed it.

I think I should have run THIS sequence after the dead disk was removed, but before the new luns were created.

mv /etc/powermt.config /etc/powermt.config.save

powermt check force

Then, after the luns were back, I would run the rest of it. The ioscans and the other powermt commands.


that's it.