Known Issues and Bugs

The following known issues and bugs affect the operation of the Oracle Solaris Cluster Geographic Edition 3.3 5/11 release.

Depending on the Order in Which Zone Cluster Setup and Geographic Edition Software Installation Is Done, Some Files Might Not Get Copied to the Zone (6968002)

Problem Summary: If you install Geographic Edition after a zone cluster is created and on a cluster that is booted into noncluster mode, not all files that are needed for Geographic Edition to work in zone clusters are copied to the zone-cluster nodes.

Workaround: Perform the following task on each node of the zone cluster:

1. Copy the following files from the /etc/cluster/geocmass/ directory to the /etc/cacao/instances/default/modules/ directory.

   com.sun.cluster.agent.cluster.xml 
   com.sun.cluster.agent.config_access.xml 
   com.sun.cluster.agent.event.xml 
   com.sun.cluster.agent.failovercontrol.xml 
   com.sun.cluster.agent.logquery.xml 
   com.sun.cluster.agent.node.xml 
   com.sun.cluster.agent.rgm.xml 
   com.sun.cluster.agent.devicegroup.xml 

2. Restart the common agent container.

   # /usr/sbin/cacaoadm restart

geo-failovercontrol Resource Fails to Start on Second Node if Connection Lost to Primary Node (6932775)

Problem Summary: When many protection groups are configured, failure of the primary node of a cluster might cause the startup of the Geographic Edition infrastructure to time out while starting on another node.

Workaround: Increase the START_TIMEOUT property of the geo-failovercontrol resource from the default of 600 seconds. This enables the geo-infrastructure resource group to fail over successfully. The required value might need to be determined by calculation, depending on the number of protection groups that are configured.