Algorithm for self-optimising workload management in a cluster
Original Publication Date: 2003-Dec-10
Included in the Prior Art Database: 2003-Dec-10
This article describes an algorithm that uses knowledge local to a routing application to dynamically select a routing strategy to optimally balance a workload within a message-queueing cluster, by allocating spare resources effectively. The algorithm avoids having to query the state of destination systems and hence removes the delay and possible instability in routing choice inherent in a closed-loop routing system.