I am using Drools Planner for an optimization problem and facing weird memory corruption issues. The problem facts and planning entities are JPA entities which have lots of relationships and are large in size. When I use a few of the entities, then the solving works ok. However when I use large number of these entities, then I get strange exceptions indicating memory corruption in the result. E.g. the serialized best solution (serialized by me for persistence purposes) appears to be have invalid class types in members. I see that there are the following exceptions in the log (for both small and large data sets) - Exception in thread "Thread-106" java.lang.RuntimeException: Unknown accessor type: org.mvel2.optimizers.impl.refl.collection.ArrayCreator@e7d1626d [2/12/13 12:35:34:309 IST] 00000174 SystemErr R at org.drools.rule.constraint.ConditionAnalyzer.analyzeAccessor(ConditionAnalyzer.java:264) [2/12/13 12:35:34:309 IST] 00000174 SystemErr R at org.drools.rule.constraint.ConditionAnalyzer.analyzeNode(ConditionAnalyzer.java:250) [2/12/13 12:35:34:309 IST] 00000174 SystemErr R at org.drools.rule.constraint.ConditionAnalyzer.statementToExpression(ConditionAnalyzer.java:467) [2/12/13 12:35:34:309 IST] 00000174 SystemErr R at org.drools.rule.constraint.ConditionAnalyzer.readInvocationParams(ConditionAnalyzer.java:457) [2/12/13 12:35:34:309 IST] 00000174 SystemErr R at org.drools.rule.constraint.ConditionAnalyzer.analyzeAccessor(ConditionAnalyzer.java:345) [2/12/13 12:35:34:309 IST] 00000174 SystemErr R at org.drools.rule.constraint.ConditionAnalyzer.analyzeExpressionNode(ConditionAnalyzer.java:326) [2/12/13 12:35:34:309 IST] 00000174 SystemErr R at org.drools.rule.constraint.ConditionAnalyzer.analyzeAccessor(ConditionAnalyzer.java:286) [2/12/13 12:35:34:309 IST] 00000174 SystemErr R at org.drools.rule.constraint.ConditionAnalyzer.analyzeNode(ConditionAnalyzer.java:250) [2/12/13 12:35:34:309 IST] 00000174 SystemErr R at org.drools.rule.constraint.ConditionAnalyzer.analyzeSingleCondition(ConditionAnalyzer.java:133) [2/12/13 12:35:34:309 IST] 00000174 SystemErr R at org.drools.rule.constraint.ConditionAnalyzer.analyzeCondition(ConditionAnalyzer.java:99) [2/12/13 12:35:34:309 IST] 00000174 SystemErr R at org.drools.rule.constraint.ConditionAnalyzer.analyzeCondition(ConditionAnalyzer.java:70) [2/12/13 12:35:34:309 IST] 00000174 SystemErr R at org.drools.rule.constraint.MvelConditionEvaluator.getAnalyzedCondition(MvelConditionEvaluator.java:83) [2/12/13 12:35:34:309 IST] 00000174 SystemErr R at org.drools.rule.constraint.MvelConstraint.executeJitting(MvelConstraint.java:270) [2/12/13 12:35:34:309 IST] 00000174 SystemErr R at org.drools.rule.constraint.MvelConstraint.access$200(MvelConstraint.java:51) [2/12/13 12:35:34:309 IST] 00000174 SystemErr R at org.drools.rule.constraint.MvelConstraint$ConditionJitter.run(MvelConstraint.java:250) [2/12/13 12:35:34:309 IST] 00000174 SystemErr R at java.util.concurrent.ThreadPoolExecutor$Worker.runTask(ThreadPoolExecutor.java:897) [2/12/13 12:35:34:309 IST] 00000174 SystemErr R at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:919) [2/12/13 12:35:34:309 IST] 00000174 SystemErr R at java.lang.Thread.run(Thread.java:770) Are these exceptions (which appear benign in the small data set solution) causing my issue? Are there any further troubleshooting that I can perform to isolate the root cause? Any help will be much appreciated. Thanks -- View this message in context: http://drools.46999.n3.nabble.com/Drools-Planner-5-5-0-Final-Memory-corru... Sent from the Drools: User forum mailing list archive at Nabble.com. _______________________________________________ rules-users mailing list rules-users(a)lists.jboss.org https://lists.jboss.org/mailman/listinfo/rules-users

> Can you try 5.5.1-SNAPSHOT? I tried with 5.5.1-SNAPSHOT from https://repository.jboss.org/nexus/content/groups/public The problem still exists > If you can reproduce with 5.5.1-SNAPSHOT, please file a jira with the > stacktrace and if possible a way to reproduce. > https://issues.jboss.org/browse/DROOLS I am trying to package the issue into a reproducible test case for the JIRA.

I have tried different combinations of large and small data and have the following observations - 1. The data corruption happens when a limit on the number of problem facts is crossed. This leads me to believe that insertion of a large amount of data in the working memory is causing this issue. 2. I am running the planner within a server side library. Each time i invoke the optimization routine and suffer this error, the server process memory keeps increasing. This indicates that memory is not being released (corruption?)

-- View this message in context: http://drools.46999.n3.nabble.com/Drools-Planner-5-5-0-Final-Memory-corru... Sent from the Drools: User forum mailing list archive at Nabble.com. _______________________________________________ rules-users mailing list rules-users(a)lists.jboss.org https://lists.jboss.org/mailman/listinfo/rules-users

The problem is fixed now. I am summarizing the resolution to provide closure to this post.