Siebel Analytics Server Administration Guide > Clustering Siebel Analytics Servers >

Performance Considerations

This section describes characteristics of the Cluster Server feature that may influence the performance of clustered Analytics Servers. You should consider these points when implementing the Cluster Server feature.

• Sessions are assigned to an Analytics Server when the session is established. A session is assigned to the server with the fewest sessions. As a result, the load on Analytics Servers can vary and Analytics Servers brought online into an operational cluster may not receive work for some time.
• Because each Analytics Server maintains its own local query results cache, back-end databases may receive the same query from multiple Analytics Servers even though the result is cached.
• Because each Analytics Server maintains its own local query results cache, Analytics Servers that are brought online into an operational cluster may respond to queries more slowly while their local cache is being populated.
• Because each Analytics Server has an independent copy of each repository and hence its own back-end connection pools, back-end databases may experience as many as N*M connections, where N is the number of active servers in the cluster and M is the maximum sessions allowed in the connection pool of a single repository. Therefore, it may be appropriate to reduce the maximum number of sessions configured in session pools.