HACMP switch over

Hi

I had an active passive cluster. Node A went down and all resource groups moved to Node B.
Now we brought up Node A. What is the procedure to bring everything back to Node A.

Node A [/] #lssrc -a | grep cl
 clcomdES         clcomdES         323782       active
 clstrmgrES       cluster          430318       active
 nimsh            nimclient                     inoperative
 gsclvmd                                        inoperative

Node B [/] #lssrc -a | grep cl
 clcomdES         clcomdES         430292       active
 clstrmgrES       cluster          360632       active
 gsclvmd                           381042       active
 clinfoES         cluster          303340       active
 nimsh            nimclient                     inoperative

Please assist. Thanks

This can be done with CSPOC easily.

# smitty hacmp
    -> System Management (C-SPOC)
         -> HACMP Resource Group and Application Management
              -> Move a Resource Group to Another Node / Site

Select one RG at a time and move it to the other node.

Hi
If I start the cluster now in node A using "smitty clstart", wont the resource groups come to node A from node B

I don't want to appear rude or uninformative, but: clusters are a really sensible class of systems and most likely the applications running on it are mission critical - this is usually the reason one decides to (undergo the trouble to) use an HA-system in first line.

Are you sure - really, really sure - you should administrate such a system with such an obvious lack of expertise? Moving a resource group is quite a basic task to do. Don't you think it would be wiser to hire an expert to work on a perhaps quite sensible system while you get some education (not some answers here - i am talking classes and/or homework)?

There are some so-called "red books" at IBM Redbooks you might want to download and study (redbooks are freely downloadable PDFs) dealing with HACMP/PowerHA. Here is a (non-complete) list of redbooks you might find interesting:

Implementing High Availability Cluster Multi-Processing (HACMP) Cookbook, SG24-6769-00

HACMP 5.3, Dynamic LPAR, and Virtualization, REDP-4027-00

IBM eServer pSeries HACMP V5.x Certification Study Guide Update, SG24-6375-01

Having said this: Move the resource group using the SMIT menu (its the easiest way) by entering

smitty cl_resgrp_move.node_site

and follow the menu there, it is self-explanatory.

I hope this helps.

bakunin

PS: i see that shockneck already answered while i was typing - the SMIT menu he pointed you to is the same one i told you the fastpath for.

If the Ressource Groups fails back automatically depends on your configuration. You can check the fallback policy in the smit menu

or e.g. with

Hi

The daemons gsclvmd and nimsh are still showing up as inoperative in node A

#lssrc -a | grep cl
 clcomdES         clcomdES         323782       active
 clstrmgrES       cluster          430318       active
 nimsh            nimclient                     inoperative
 gsclvmd                                        inoperative

So will the resource group movement work fine?

You don't need nimsh.
Are you using concurrent vg's?

Yes. The VG is of the type concurrent VG

gsclvmd should be started with HACMP services.
Have you already done smit clstart?

Thanks for your reply friends...
I have started the cluster services via smitty clstart, and then moved the resource groups via c-spoc.
Everything worked fine as expected