How is the transaction state info is maintained in a failover scenario.


The repositoy serever takes care of maintaining the transaction state in case of failover


We have to use clustering in our environment.

The user for Application is a Http client , they
submit requests using JMS bridge to MQ and through
MQ it is picked up by our Application. So is it possible to go for
clustering in this case, as even though the client is Http
client, requests come through MQ.

Another user of Application submits requests using webMethods
File Polling. So again we are not sure whether Server Clustering
is possible in this scenario.
Please suggest.