Prabakar, The alternative would be a custom algorithm, which will be slower than Planner. However, you should limit how long you want Planner to solve. I'd recommend a short time in your case, such as 30 seconds or 1 minute. It depends on the size of your data set of course. Planner 5.3 will support out-of-the-box construction heuristics. If time is really short, you might you use a construction heuristic and no tabu search or SA at all (although I 'd still recommend to do a few seconds of TS or SA anyway).
Also, you probably don't want more Planners running at the same time than your numberOfCpuCores - 1. In that situation you 'll want to queue the planning jobs: there's no jbpm-planner integration yet that does that for you yet, but patches (pull requests) are welcome ;) Op 14-07-11 21:39, DroolersEye schreef: > Hi Geoffrey De Smet, > > thanks lot for your reply, > > I almost try all the examples in the examples GUI main page. Now I > understand why it takes long time...to get the best out of best (I spend > some time to understand the algoritms like tabu search, simulated > annealing). > Here I have the suituation like this, In my process flow (using JBPM5), one > of the node providing input(pojo) to the Drools Planner and the process > waiting for the result from from the Drools Planner(pojo) that output pojo > is the input data for the next node in the process flow. > In the above scenario what is your advise to me, Is Drools Planner is good > in this flow? I worried for the time taken by the planner to complete the > task. This will affect the whole process time. > > Please advise me, > > > ----- > > with kind regards, > Prabakar > -- > View this message in context: > http://drools.46999.n3.nabble.com/Re-rules-users-Drools-Planner-examples-applications-running-more-than-10-minutes-tp3168326p3170133.html > Sent from the Drools: User forum mailing list archive at Nabble.com. > _______________________________________________ > rules-users mailing list > [email protected] > https://lists.jboss.org/mailman/listinfo/rules-users > -- With kind regards, Geoffrey De Smet _______________________________________________ rules-users mailing list [email protected] https://lists.jboss.org/mailman/listinfo/rules-users
