8.3.1 Overview: Clustering the Sametime 8.5 Meeting Server 

Master Table of Contents | Chapter Table of Contents | Previous | Next


The Lotus Sametime 8.5 Meeting server cluster provides failover in terms of high availability.

Sametime Meeting Server Clustering Highlights:

• Sametime leverages WebSphere Unified Clustering Framework (UCF). UCF monitors node health and notifies applications when a clustered server is down.
• Sametime 8.5 meeting server includes the “WebSphere Proxy” with special Filters used for routing HTTP request to the appropriate cluster member.
• WebSphere Network Deployment (ND) is installed “standard”.
• All meetings are assigned to a node of the cluster and only transition in case of a failure.
• Meeting State is periodically saved to database to provide application failover.

The deployment in this example contains a two Meeting servers which are clustered. A cluster of Lotus Sametime servers requires at least one WebSphere proxy server to handle routing and caching tasks. When you install Lotus Sametime on a node in the cluster, WebSphere Application Server and WebSphere proxy server are also installed. The WebSphere proxy server merely needs to be configured for use. To reduce the resource load on product nodes and avoid port conflicts, you may choose to install a stand-alone WebSphere proxy server on a separate computer instead of using the instance that was installed on a Lotus Sametime node. We have done this in the RedWiki example.


User experience of a clustered meeting example:

Several users are in a meeting. The Meeting server has been stopped for some reason. The client realizes that it can't communicate to the original server any longer and issues a dialog box to the user asking if they would like to retry the connection to the meeting.



After clicking the Try Again button, the client rejoins the meeting which has failed over the other cluster member. The entire process takes less than 30 seconds.