Serviceguard Extension for RAC Version A.11.18 Release Notes, 4th Edition, September 2008

NOTE: All nodes in the cluster must be either all SGeRAC nodes or Serviceguard nodes. SGeRAC
cluster must have SGeRAC software installed on all nodes.
1. Halt the cluster.
2. Use swinstall to install SGeRAC on each node.
3. Modify the cluster configuration ASCII file as necessary for RAC use, add
OPS_VOLUME_GROUP if required, then run the cmapplyconf command.
4. Reboot each node.
5. Restart the cluster.
Removing SGeRAC
To remove your software, run the SD-UX swremove command.
NOTE: If you are removing SGeRAC and wish to continue using the cluster as an Serviceguard
cluster, there are slightly different steps. See section “Removing Serviceguard Extension for RAC
Without Removing Serviceguard”.
Use the following steps to remove the software from each node:
Online Node Reconfiguration:
1. Halt the cluster services on the node(s) to be removed. Edit the cluster ASCII file to remove
the nodes from which you removed SGeRAC, then re-apply the configuration. The cluster
remains up and running.
2. Issue the swremove command to remove the software.
Offline Node Reconfiguration:
1. Halt the cluster and remove the node from the configuration. If you intend to continue
running the SGeRAC cluster on the remaining nodes, edit the cluster ASCII file to remove
the nodes from which you removed SGeRAC, then re-apply the configuration.
2. Optionally, restart the cluster on the remaining nodes.
3. Issue the swremove command to remove the software. If SGeRAC is being deinstalled from
more than one system, it should be removed from one system at a time.
Optionally, after de-installing SGeRAC, you may use the swremove command to de-install
Serviceguard.
Removing Serviceguard Extension for RAC Without Removing Serviceguard
If you want to remove SGeRAC and return the cluster to use as a Serviceguard cluster, use the
following steps:
1. Halt the cluster.
2. Use swremove to remove SGeRAC from each node.
3. Reboot each node.
4. Edit the cluster configuration ASCII file to remove any OPS_VOLUME_GROUP entries, then
run the cmapplyconf command.
5. Restart the cluster.
Troubleshooting Tips
The core dump locations for the cluster manager and shared logical volume manager daemons
are in /var/adm/cmcluster and /etc/lvmconf respectively. For cmgmsd, the core dump
location is in /var/adm/cmcluster/cmsgmsd. By default, log messages from these daemons
appear in /var/adm/syslog/syslog.log.
18 Serviceguard Extension for RAC Version A.11.18 Release Notes