HP Integrity rx5670 Windows Integrity Cluster Installation and Configuration G - Page 36

Setup, configuration, validation, and maintenance of the cluster Upgrading individual nodes in the future Upgrading individual nodes in the future At some point after your initial installation and configuration you will want to upgrade the software and drivers installed on each node in your cluster, as well as add the latest system updates and security fixes. This is an inevitable system administration chore that must be done regularly if you want to keep your Integrity servers up-to-date and secure. With clustered systems you no longer have to do all of your maintenance during those rare times when all of your users are off-line. Instead, you simply wait until a convenient, off-peak time when one of the nodes in the cluster can be taken off-line for maintenance and its workload distributed among the remaining nodes. Before the upgrade, however, you must evaluate the entire system to verify that the remaining nodes can handle the increased capacity (while the node being upgraded is taken off-line). Basically, you pick the node you want to upgrade, then use the Cluster Administrator to move all of the clustered resources onto one or more of the remaining nodes. This can also be accomplished using scripts. Once all of the resources have been failed-over to the other nodes, the first node is free and ready for upgrading. For more information about how to upgrade your Integrity servers with the latest drivers and QFEs, see the latest Smart Setup Guide for Integrity servers at http://docs.hp.com/en/windows.html, under the section, "Windows 64-bit on HP Integrity Servers". Once the upgrade to the first node is complete, reboot it if necessary and move the resources back to it. As soon as possible you should repeat this process to upgrade the other nodes in the cluster. This minimizes the amount of time the nodes are operating with different versions of software or drivers. 36 Chapter 2