Legato Co-StandbyServer 2000

Use this software to create a fail- over cluster 

Thursday, September 12, 2002

This software creates a fail-over cluster using two Windows 2000 servers. Under normal circumstances, one of these machines acts as the primary server providing access to applications and resources for the clients. If due to some reason, the primary server goes down, the second server automatically takes over the primary server’s role. This ensures that network resources are always available to the clients. This makes it suitable for applications where you can’t afford to have any downtime.

Some of the resources that can be managed using Co-StandbyServer are data, applications, IP addresses, services, server aliases, storage devices, and printer queues. If any of these resources fail on the primary server, they can be made available on the secondary server. Both servers must have two network cards each. One pair acts as a dedicated link between the two servers, while the other connects to the main network. Both the links were 10 Mbps full duplex. The best part is that the cluster machines don’t need to have similar configurations.

We tested the server by mirroring a partition on the two servers. Initially, the server takes a while to mirror the complete data between the two partitions. While the actual server configuration is not very difficult to do, it took us a while to have it up and running, because of the documentation. It covers the prerequisites, installation process, and descriptions of the various menu items. However, a step-by-step example illustrating its functions would definitely make things simpler.

Once the cluster was up and running, it worked like a charm. We tried a number of tests on it right from a manual fail-over to an automatic one from one server to another. We pulled out the network cable of the primary server, and it took about 22 seconds to show the secondary server take complete control. We also switched off the primary server, and the failover to the second server happened automatically. Another interesting feature is that the failover is bi-directional, meaning control can be shifted from any server to any server in the cluster.

The Bottom Line: The beauty of the software is that you don’t need machines with the same hardware to build the cluster. It does automatic failover to the second server in case the primary server fails. Overall, a useful software for high-availability application requirements.

Anil Chopra and Santhosh Unny at PCQ Labs

Page(s)   1