[jboss-jira] [JBoss JIRA] (DROOLS-1441) Nested accumulate makes DRL twice as slow in statefull session

[Geoffrey De Smet (JIRA)]

     [ https://issues.jboss.org/browse/DROOLS-1441?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Geoffrey De Smet updated DROOLS-1441:
-------------------------------------
    Description: 
To reproduce: comment out to replace the 2 rules in this DRL:
  https://github.com/droolsjbpm/optaplanner/blob/master/optaplanner-examples/src/main/resources/org/optaplanner/examples/tennis/solver/tennisScoreRules.drl
and run TennisApp

In the sampler profiling, see attachement 350, the original way does setObject() 60% of the time and determineIdentityHashcode 20%, but in the new one way it's vica versa.
so the number of setObjects() calls is probably the same, it's just the determineIdentityHashCode() that blow up.

  was:
To reproduce: comment out to replace the 2 rules in this DRL:
  https://github.com/droolsjbpm/optaplanner/blob/master/optaplanner-examples/src/main/resources/org/optaplanner/examples/tennis/solver/tennisScoreRules.drl
and run TennisApp



> Nested accumulate makes DRL twice as slow in statefull session
> --------------------------------------------------------------
>
>                 Key: DROOLS-1441
>                 URL: https://issues.jboss.org/browse/DROOLS-1441
>             Project: Drools
>          Issue Type: Bug
>          Components: core engine
>    Affects Versions: 7.0.0.Beta6
>            Reporter: Geoffrey De Smet
>            Assignee: Mario Fusco
>         Attachments: Selection_350.png
>
>
> To reproduce: comment out to replace the 2 rules in this DRL:
>   https://github.com/droolsjbpm/optaplanner/blob/master/optaplanner-examples/src/main/resources/org/optaplanner/examples/tennis/solver/tennisScoreRules.drl
> and run TennisApp
> In the sampler profiling, see attachement 350, the original way does setObject() 60% of the time and determineIdentityHashcode 20%, but in the new one way it's vica versa.
> so the number of setObjects() calls is probably the same, it's just the determineIdentityHashCode() that blow up.



--
This message was sent by Atlassian JIRA
(v7.2.3#72005)