Browse Prior Art Database

Dynamic message-routing algorithm in a cluster

IP.com Disclosure Number: IPCOM000020706D
Original Publication Date: 2003-Dec-10
Included in the Prior Art Database: 2003-Dec-10
Document File: 2 page(s) / 78K

Publishing Venue

IBM

Abstract

This article discloses an autonomic algorithm for routing messages according to what processing a message requires and what destinations are available that can provide that processing within a message-queuing cluster. Processing is described in a standard format such as XML, and routing applications subscribe to the descriptions of processing carried out at remote destinations. Hence changes in processing on a remote machine causes the routing strategy to be altered dynamically.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 52% of the total text.

Page 1 of 2

Dynamic message-routing algorithm in a cluster

At present, a typical routing application in a message-queueing system would require a hard-coded list of destintions and corresponding connection parameters, as well as rules to determine which message should be sent to which destinations and what to do if a particular message is unavailable. However, in an idea autonomic system, we would want messages routed according to a policy set by system adminstrators at run time without having to recode the rules and connection parameters . A better method would be to be able to route messages according to what processing the message requires. The algorithm described here does this. The main advantage is less system adminstration effort since the algorithm determines the required destination itself, and can query the connection parameters required in real time. Also, this algorithm allows changes to the routing strategy to be made more quickly and easily, since changing the application that is triggered by and gets messages for processing from a particular destination queue will actually change which messages are sent to that destination queue.

    The algorithm enables autonomic message routing in that it monitors destination queues for changes to the processing that is done to messages received on that queue, analyses any changes to the destination queue and uses the information gathered to plan what messages should be routed to particular destinations, and then executes its routing strategy by querying individual messages for what processing they require next in the workflow. An embodiment using IBM's WebSphere MQ product and the clustering functionality therein, would work in the following manner:

    Cluster queues may have a textual description. Hence, it is possible to build an application which will inquire a repository broker to get the text descriptions of all cluster queues (which would preferably be in XML format). The text description could detail the processing carried out by the particular application triggered by that queue
i.e. what would happe...