Configuring resource groups in HACMP

You can perform the steps described in this section from the command line or through the SMIT configuration utility. See your IBM documentation for information about starting and using the SMIT utility. This document explains how to use SMIT to configure resource groups. The steps below assume you are using HACMP5 2. If you are using a later version, see your operating system documentation. See the IBM HACMP for AIX documentation for information about configuring resource groups from the command line.

Shut down the cluster services on both nodes in graceful mode, then log in to start IP addresses of the primary node as “root” and perform these tasks:

Start SMIT from the command line - smit hacmp.
From Extended Configuration menu:
- Select Extended Resource Configuration.
- Select HACMP Extended Resource Group Configuration.
- Select Add a Resource Group if you are creating a new resources group, or select Change/Show a Resource Group if you are changing an existing resources group.

Select the appropriate start, fallover, and fallback policies when defining the resource groups:

For the Resource Group for the primary companion:

Field name	Enter
Resource Group Name	[<resgrp1>]
Participating Nodes (Default Node Priority)	[<primary_node> <secondary_node>]
Start Policy	“Online On Home Node Only”
Fallover Policy	“Fallover To Next Priority Node In The List”
Fallback Policy	“Fallback To Higher Priority Node In The List”

Click OK.

For the Resource Group for the secondary companion, Asymmetric Failover Configuration:

Field name	Enter
Resource Group Name	[<resgrp2>]
Participating Nodes (Default Node Priority)	[<secondary_node>]
Start Policy	“Online On Home Node Only”
Fallover Policy	“Fallover To Next Priority Node In The List”
Fallback Policy	“Fallback To Higher Priority Node In The List”

Click OK.

For the Resource Group for the secondary companion, as a Symmetric Failover Configuration:

Field name	Enter
Resource Group Name	[<resgrp2>]
Participating Nodes (Default Node Priority)	[<secondary_node> <primary_node>]
Start Policy	“Online On Home Node Only”
Fallover Policy	“Fallover To Next Priority Node In The List”
Fallback Policy	“Fallback To Higher Priority Node In The List”

Click OK.

The start and stop policies can also be “Bring Offline (On Error Node Only)” and “Never Fallback” for secondary companion in Asymmetric HA configuration.

Define the primary and secondary companions as application servers in HACMP. From Extended Configuration menu:

Select Extended Resource Configuration
Select HACMP Extended Resources Configuration
Select Configure HACMP Applications
Select Configure HACMP Application Servers

Select Add an Application Server for defining a new Application Server or Change/Show an Application Server if you are changing an existing Application Server, then enter these values:

Field name	Enter
Server Name	the corresponding companion server name for this Application Server being created.
Start Scrip	the full path name of the corresponding script you created in section “Modifying the ASE_HA.sh script”, for this companion server. Enter “monitor” as argument for the start script.
Stop Script	the full path name of the corresponding script you created in section “Modifying the ASE_HA.sh script”, for this companion server. Enter “failover” as argument for the stop script corresponding to primary and symmetric secondary companion servers. Enter “stop” as argument for asymmetric secondary companion server.
Application Monitor Name(s)	leave blank

Click OK when information is completed.

As examples,

For the primary companion server MONEY1:

Field name		Enter
	Server Name	[MONEY1].
	Start Scrip	`[/usr/sbin/cluster/events/RUN_MONEY1.sh monitor]`
	Stop Script	`[/usr/sbin/cluster/events/RUN_MONEY1.sh failover]`
	Application Monitor Name(s)	[ ]
For secondary companion server PERSONNEL1 in Symmetric HA configuration:
	Server Name	[PERSONNEL1]
	Start Script	`[/usr/sbin/cluster/events/RUN_PERSONNEL1.sh monitor]`
	Stop Script	`[/usr/sbin/cluster/events/RUN_PERSONNEL1.sh failover]`
	Application Monitor Name(s)	[ ]
For secondary companion server PERSONNEL1 in Asymmetric HA configuration:
	Server Name	[PERSONNEL1]
	Start Script	`[/usr/sbin/cluster/events/RUN_PERSONNEL1.sh monitor]`
	Stop Script	`[/usr/sbin/cluster/events/RUN_PERSONNEL1.sh stop]`
	Application Monitor Name(s)	[ ]

Only the secondary node is listed in the Participating Nodes field for secondary resource group in Asymmetric HA configuration, whereas both nodes are listed in reverse order, when compared to primary resource group, in the Symmetric HA configuration.

Configure each of the resource groups you defined for the Application Servers. From Extended Configuration menu:
- Select Extended Resource Configuration.
- Select HACMP Extended Resource Group Configuration.
- Select Changes/Show Resources and Attributes for a Resource Group.
- On the Single Select List, select the resource group you want to configure.
- In the Application Server field, enter corresponding application server name defined in step (4) for the resource group (such as, primary companion name for primary resource group and secondary companion server name for the secondary resource group).
  
  Enter information in all the required fields, such as IP Label, Volume Groups, and File systems. Repeat this step for each of the companions.
Synchronize the cluster resources. Using smit on the node where you performed steps 1 through 5, go to Extended configuration Screen and select “Extended Verification and Synchronization”. This propagates the changes you made to all the other nodes within the same cluster.

In some cases, you may need to stop the HACMP services and restart both nodes before performing the synchronization. Verify that the synchronization does not produce any errors.