High-Volume Web Sites Team - More about High-Volume Web Sites

Our testing verified that heavily loaded WebSphere Portal clusters are resilient in the face of failures in the cluster infrastructure - including failures of network adapters or failures in which application server clones become unavailable for whatever reason.

Workload is successfully redistributed to the surviving application server clones when a failure occurs, and also is successfully redistributed when the inaccessible clones become available again.

We established a number of best practices parameter settings that help to minimize the impact of infrastructure failures on the users of the portal.