7x24 maintenance
The goal of high availability is to keep systems up and running, allowing continuous access to critical applications. In many enterprises, it has become necessary to keep applications running seven days a week, 24 hours a day. With proper planning, customizing, and monitoring, a PowerHA® SystemMirror® cluster can provide nearly continuous availability, interrupted only by scheduled, necessary maintenance.
These topics are a collection of information describing the issues and procedures involved in keeping a cluster running as closely as possible to a 7 X 24 basis.
Throughout all stages of cluster administration - planning, configuration, maintenance, troubleshooting, and upgrading - here are tasks you can do and systems you can put in place to help ensure your cluster's nearly continuous availability.
Once you have configured the cluster and brought it online, it is very important to do maintenance tasks in as non-disruptive a way as possible. The PowerHA SystemMirror cluster is a distributed operating system environment. Therefore, maintaining a PowerHA SystemMirror cluster requires attention to some issues that have different ramifications in the cluster environment compared to maintaining a single-server system.
Making changes to a cluster must be thoroughly planned, since changes to one component may have cascading effects. Changes on one node may affect other nodes, but this may not be apparent until fallover occurs (or cannot occur due to a non-synchronized change to the cluster). Some of the do's and don'ts of cluster maintenance are explained in this topic.
Setting up and following regular preventive maintenance procedures helps alert you to any potential problems before they occur. Then you can take timely action or plan fallovers or cluster downtime at your convenience as necessary to deal with any impending issues.