We want to cluster two of our Red Hat AS 2.1 Web servers and configure the disk array so it can be shared between the servers (so both can see the same drives at the same time). Where can we get more details about clustering?
Check the Red Hat Cluster Suite home page. The suite is available for Enterprise Linux Version 3 as an annual subscription service, and can provide high-availability failover capabilities and IP load balancing. It is based on the Kimberlite open source clustering system.
High-availability services can be run on only one cluster member server at a time, but you can run separate services on each cluster member while sharing visibility into the same disk array if you use a dual-controller hardware RAID disk array. Each member in a two-member cluster periodically writes time stamp and cluster-state information to two shared cluster partitions located on shared disk storage. If a member cannot write to both the primary and shadow shared cluster partitions at start-up, it can’t join the cluster. Also, if a member doesn’t update its time stamp and heartbeats to the system fail, it is removed from the cluster.
