Print-friendly Version

How to Stop Monitoring a Heavily Used Resource in a Sun Cluster 3.x Environment

Karan Grover, December 2007

This tech tip provides information about how to stop monitoring a heavily used resource in a Sun Cluster 3.x resource group when you see a high load on that particular resource. This scenario can be seen, for instance, when batch jobs are running and pushing database applications to the maximum extent, which causes the databases to "go numb" and the monitors to initiate a failover.

In order to stop the resource group from failing over to other nodes and causing a loss in connectivity, you can stop the monitoring on the resource for the duration of the high load by using the following procedure. If the high load occurs predictably at known times, you could even automate the process by using cron.

The commands to disable the monitoring for a resource in a particular resource group are as follows:

1. Run the following command to stop monitoring the resource:

scswitch -M -n -j <resource-name>

Note: You need solaris.cluster.resource.admin RBAC authorization to use the scswitch command with the -e or -n option. See the scswitch(1m) man page for more information.

For example, if the listener of a particular Oracle database instance is initiating a failover, you can stop the monitoring on the initiator resource as follows, where <oralisten-serv-res> is the name of the resource you are trying to disable:

scswitch -M -n -j <oralisten-serv-res>

Note: I tried this on Oracle 9.2.0.5.0, but it should work with Oracle 9i or any version of Oracle.

2. To re-enable monitoring on the resource after the load period is over, use the following command:

scswitch -M -e -j <resource-name>

This command initiates monitoring on the resource, so that in the future, a failover can be initiated when necessary.

Unless otherwise licensed, code in all technical manuals herein (including articles, FAQs, samples) is provided under this License.