Originally posted by: SystemAdmin

Hi

I have a HACMP 5.3 on AIX 5.3 ML05 connected to FastT 4800. This is a two node cluster and with concurrent resource group(active-active).

I want to know, how to add a LUN from the FastT to the concurrent VG?
Can I use the extendvg command to do so and followed by synchronizing the cluster? Is this safe?

As it been stated in FastT guide that FastT disk cannot be added with cspoc, what may be the other options?

Please advice.

thanks
Kuntal

Originally posted by: alethad

I'm not familiar with your Fast T but if you have your cluster RG configured to volume groups there should be no problem. It doesn't matter what disks are in the volume group HA grabs all the disks in the volume group. We have shark disks and we add new ones thru the OS as normal procedure. I have the VG's configured in the RG with no filesystems or disks listed. So I add disks and make sure they are seen by both nodes or whatever then I just run verify & sync afterward just to be sure. No topolgy really changes when you add a disk that belongs to a VG that is already included in the RG.
If you're adding a new volume group you can do it all thru the OS then add it to the RG config just as you did with the existing VG's.
It really depends on how you have your cluster configured.

You'll want to verify & sync afterward no matter what then schedule a failover test with your new disks.

Good luck.

Originally posted by: SystemAdmin

The best way is to use the CSPOC (Cluster Single Point Of Control) to extend the VG size..It will automatically updates the ODM of both the nodes..hence no need for synchronization...

Hope this will help

Originally posted by: alethad

HACMP 5.3 on my system does not allow you to extend volume groups. Check yours Kuntal just to be sure. You can add or change them. If yours is the same as mine then you'll have to do it thru the regular SMIT screens. You may have even have to re-import it on the other nodes. But run your verify first and see.
Good luck.

Originally posted by: SystemAdmin

hi

Yes that's what happening to me too. That is FastT storage ( DS4800 ). Recently I have come across a FastT document saying that CSOPC does not support addition of FastT disks to the cluster. But not sure it is due to this or a HACMP 5.3 problem. So everytime , a new PV needs to be added to the cluster VGs, I need to do it by bringing down the cluster and reimporting the VGs. Varification fails If I dont do this.I have come across clusters having sharks and ESS storage attach and they don't have this problem.
Is there any APAR or any kind of fix for HACMP 5.3 that may help ?

Originally posted by: alethad

There's no fix for the absence of extending the VG's that's just how HACMP 5.3 was made. You can add disks thru CSPOC but I don't use it.

You can also use the importvg with the -L option which is sort of a lazy update command. You varyon the VG on the primary then run importvg -L on the secondary node and that will update the VG on the other node. Are you familisr with that? I use it occasionally. It doesn't require downtime to do it. You'd want to use it cautiously.

alethad