Configure NNMi for High Availability

The two distinct phases of configuring NNMi for High Availability (HA) are as follows:

  1. Copy the NNMi data files to the shared disk.

  2. Configure NNMi to run under HA.

Designate one HA cluster node as the primary NNMi management server. This is the node you expect to be active most of the time. Configure the primary node, and then configure all other nodes in the HA cluster as secondary nodes.

Caution You cannot configure NNMi for HA simultaneously on multiple cluster nodes. After the HA configuration process is completed on one cluster node, proceed with the HA configuration on the next node, and so forth until NNMi is configured for HA on all nodes in the cluster environment.

  During failover, the NNMi console is unresponsive. After failover completes, NNMi users must log on to continue their NNMi console sessions.

Note If you encounter errors during HA configuration, do the following:

  1. Unconfigure NNMi from the HA environment by running the nnmhaunconfigure.ovpl command.
  2. Correct the condition indicated by the error message(s).
  3. Reconfigure NNMi into the HA environment by running the nnmhaconfigure.ovpl command.

    (RHCS only) For the nnmhaconfigure.ovpl and nnmhaunconfigure.ovpl commands to work properly, the <failoverdomains/> tag must exist in the /etc/cluster/cluster.conf file.

    The <failoverdomains/> tag is embedded within the resource manager section, for example:

    ...
    ...
    <rm>
     <failoverdomains/>
    </rm>

    The nnmhaconfigure.ovpl command requires the<failoverdomains/> tag to create the NNMi resource group, using the following example structure:

    ...
    <rm>
      <failoverdomains>
        <failoverdomain name="<rg-name>-dom" nofailback="0"
    ordered="0" restricted="1">
    	<failoverdomainnode name="<node1>" priority="1"/>
    	<failoverdomainnode name="<node2>" priority="1"/>          </failoverdomain>
       </failoverdomains>
       <service autostart="1" domain="<rg-name>-dom" 
    exclusive="0" name="nnmha" recovery="relocate">
          <ip address="<addr>" monitor_link="1">
              <fs device="<nnmhalvol>" force_fsck="1" 
    force_unmount="1" fsid="" fstype="ext3" 
    mountpoint="<nnm-hamount>" name="nnmha-mount" 
    options="" self_fence="0">
              <NNMscript GLOBAL_VARIABLES="NNM_INTERFACE=
    <virtual hostname>;HA_LOCALE=en_US.UTF-8;
    HA_MOUNT_POINT=/<nnm-hamount>" 
    file="/var/opt/OV/hacluster/<rg-name>/nnmharhcs" 
    name="nnmha-APP"/>
              </fs>
         </ip>
      </service>
    </rm>

    The nnmhaunconfigure.ovpl command also requires the above structure to remove the node's failoverdomain entry.

    For more information, see the nnmhaunconfigure.ovpl and nnmhaconfigure.ovpl reference pages, or the Linux manpages.