Tek-Tips is the largest IT community on the Internet today!

Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!

  • Congratulations gkittelson on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!

PVID's Off

Status
Not open for further replies.
Apr 12, 2001
46
US
IBM "field repair man" installed 7 new disks in an SSA cabinet. He booted the system up in Diag (I think, wasn't there) then checked all the disks. Now all the PVID's don't match up and I can't find directories that I know is there and the disks in LV's are wrong. This system is running HACMP. Lucky this is the test/fail over box. What is the best way to get the system back to its original config or am I out of luck??
 
tristan,

I think it may be worth posting up the lspv's from the production and standby machines. If the disk order is now different you could try removing the new disks on the standby box and running

cfgmgr -vl ssar

If not you have to add the disks manually through smit and select the right location code to get them in the same order.

I think that is your problem anyway, but this will not have effected your existing volume groups and disks. Post up if I have interpreted your problem incorrectly.

Cheers

PSD
IBM Certified Specialist - AIX V4.3 Systems Support
IBM Certified Specialist - AIX V4 HACMP
 
Well. When I do an lsvg -p tempvg it returns all the wrong hdisks. The hdisks it is returning now matches up to the wrong vg? Not really sure if that helps.
 
tristan,

Well that makes no sense at all. Did he replace disks? if the disks were part of a volume group then they should have retained the same hdisk numbers. I suspect that the engineer may have removed all hdisks and pdisks and re-ran config. manager on the second node thus creating a scenario similar to your problem.

I will need some outputs to offer more assistance

lspv
lsvg -p vgname (for all vg's)

Cheers

PSD
IBM Certified Specialist - AIX V4.3 Systems Support
IBM Certified Specialist - AIX V4 HACMP
 
Currenlty that node is down. I am about to call IBM. You could be right when you said that the guy removed the disks then reran cfgmgr. Like I said. I wasn't in there when he booted back up. If I did remove all the drives with rmdev -C | ssar -R the reran cfgmgr, would that reassign all my drives correctly?
 
Tristan,

In theory it would do this as it based on serial number order but you may have to do it on both machines so they go in the same order, this sometimes does not work as expected due to the cabling of the SSA adapters but it cannot be any worse.

Best of luck.

Like I said the only other way is to remove them all on the second node and add them one by one through smit.

PSD
IBM Certified Specialist - AIX V4.3 Systems Support
IBM Certified Specialist - AIX V4 HACMP
 
I removed all the devices then ran cfgmgr again. It gave new pvid's to the disks. Now, for example, hdisk42 with PVID of 123 is now associated with hdisk 44. However hdisk42 should be pointing to tempvg now is pointing to buildvg. This is very confusing!!! Who knows :( Maybe I can get on at Babbages and sell video games!
 
tristan,

It sounds like a nightmare you will have to export and import the volume groups from each system and import them in so that at least the correct disks point to the correct volume groups. I would log this with IBM, see if they can sort out the mess. I need to look at the system really to work out the problems.

Best of luck anyway.

PSD
IBM Certified Specialist - AIX V4.3 Systems Support
IBM Certified Specialist - AIX V4 HACMP
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor

Back
Top