Browse Prior Art Database

System and method of rerouting message from failure message engine under clustering messaging environment

IP.com Disclosure Number: IPCOM000241080D
Publication Date: 2015-Mar-26
Document File: 7 page(s) / 60K

Publishing Venue

The IP.com Prior Art Database

Abstract

We invent a system and method of re-routing messages from failed nodes, to other available cluster members. With this system and method, a message route agent component is introduced, which is able to access a log written by failed message node and extract messages not yet processed. It can write into a data store (can be a database or a file). A gateway message node can then route messages to other cluster members by reading messages from the store. If / when the failed message node becomes available, it communicates with message route agent to determine what messages have been processed before it performs transaction recovery, then rollback messages that have been redirected. The method will ensure message redirected from failed message engine within short time and ensure transaction consistency.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 79% of the total text.

Page 01 of 7

System and method of rerouting message from failure message engine under clustering messaging environment

In clustering message engine environment, a gateway message node is responsible to route message to cluster nodes. A message route agent is used to administrate message engine running on the same node with the invented feature enabled. Once on, it will register to the message route agent. Message route agent communicates the information with gateway node, then they will begin to send heartbeatsignals every established internals to the message node to check if it is responsive.

 When a message node turns on high availability function, it writes cluster in-bound messages in a separate log, it will also record what messages have been processed. When it receives heartbeats from gateway node and successfully respond, it begin to clean up log and rewrite. If gateway detects the message engine is not responsive, then it will inform the message route agent immediately. The message route agent then reads the log written by the now presumed "failed" message node, to identify which messages are not processed since the last acknowledged heartbeat signal sent, then write messages with message body, destination name into the Unconsumed In-bound Message (UIM) table and notify gateway node. The Gateway message engine access the UIM table and redirects message to other cluster members. When the failed node is available after a successful restart, it won't proceed on transacti...