[rules-users] Drools Flow, threads and persistence in a clustered environment

[Alexandros Karypidis]

Hello,

We plan to use Drools (Flow) in our application for managing 
long-running processes. High-availability is important to us and 
therefore we target a clustered JBoss environment with 3 nodes. We are 
looking for advice on how to approach this from a Drools perspective. 
Can anybody share their experiences on running Drools in a cluster? The 
following scenarios come to mind:

1) A "singleton" ksession runs constantly on one of the nodes and does 
all the work; it migrates when the node goes down to another machine.

2) Multiple "ksession" instances on different nodes (for distributing 
load); when one fails the other picks up its processes by having a 
second "ksession" instance created on it and resuming execution.

Are there any wrapper/utility libraries for Drools that help in 
clustered environments?

Thank you