[
https://issues.apache.org/jira/browse/JENA-1260?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15670611#comment-15670611
]
Dave Reynolds commented on JENA-1260:
-------------------------------------
The TransitiveEngine is a bit of special case because it has a (hacky)
implementation of rdfs:subPropertyOf aliases. The caches in question are only
used when you have subPropertyOf alias (which isa corner case) but adding a
triple which introduces a new alias causes those caches to get flagged for
rebuild. So it's hard to provoke the very specific situation that will be
affected without triggering a cache rebuild.
I don't think this is performance issue because it's only when you add triples
like ":p rdfs:subPropertyOf rdfs:subPropertyOf" that this cache rebuild is
triggered and those are very rare outside of test cases!
However, having thought about it more I realised that using a
TransitiveReasoner directly, rather than an RDFS reasoner, gives more control
and have generated a deterministic test case that failed before the fix and now
passes.
I'll submit that as a PR (partly as a check that I'm appropriately set up for
doing PR's for Jena).
> NullPointerException when using an inferencing model under multithreaded load
> with query and delete
> ---------------------------------------------------------------------------------------------------
>
> Key: JENA-1260
> URL: https://issues.apache.org/jira/browse/JENA-1260
> Project: Apache Jena
> Issue Type: Bug
> Components: Jena
> Affects Versions: Jena 3.0.0
> Reporter: Stephen Owens
> Assignee: Dave Reynolds
> Attachments: TransitiveEngineFailureTest.java
>
>
> {noformat}
> java.lang.NullPointerException
> at java.util.HashSet.contains(HashSet.java:203)
> at
> org.apache.jena.reasoner.transitiveReasoner.TransitiveEngine.triage(TransitiveEngine.java:194)
> at
> org.apache.jena.reasoner.transitiveReasoner.TransitiveEngine.add(TransitiveEngine.java:234)
> at
> org.apache.jena.reasoner.rulesys.FBRuleInfGraph.prepare(FBRuleInfGraph.java:431)
> at
> org.apache.jena.reasoner.BaseInfGraph.requirePrepared(BaseInfGraph.java:530)
> at
> org.apache.jena.reasoner.rulesys.FBRuleInfGraph.findWithContinuation(FBRuleInfGraph.java:557)
> at
> org.apache.jena.reasoner.rulesys.FBRuleInfGraph.graphBaseFind(FBRuleInfGraph.java:587)
> at org.apache.jena.graph.impl.GraphBase.find(GraphBase.java:255)
> at
> org.apache.jena.sparql.engine.iterator.QueryIterTriplePattern$TripleMapper.<init>(QueryIterTriplePattern.java:75)
> at
> org.apache.jena.sparql.engine.iterator.QueryIterTriplePattern.nextStage(QueryIterTriplePattern.java:49)
> at
> org.apache.jena.sparql.engine.iterator.QueryIterRepeatApply.makeNextStage(QueryIterRepeatApply.java:108)
> at
> org.apache.jena.sparql.engine.iterator.QueryIterRepeatApply.hasNextBinding(QueryIterRepeatApply.java:65)
> at
> org.apache.jena.sparql.engine.iterator.QueryIteratorBase.hasNext(QueryIteratorBase.java:111)
> at
> org.apache.jena.sparql.engine.iterator.QueryIterBlockTriples.hasNextBinding(QueryIterBlockTriples.java:63)
> at
> org.apache.jena.sparql.engine.iterator.QueryIteratorBase.hasNext(QueryIteratorBase.java:111)
> at
> org.apache.jena.sparql.engine.iterator.QueryIteratorWrapper.hasNextBinding(QueryIteratorWrapper.java:39)
> at
> org.apache.jena.sparql.engine.iterator.QueryIteratorBase.hasNext(QueryIteratorBase.java:111)
> at
> org.apache.jena.sparql.engine.iterator.QueryIteratorWrapper.hasNextBinding(QueryIteratorWrapper.java:39)
> at
> org.apache.jena.sparql.engine.iterator.QueryIteratorBase.hasNext(QueryIteratorBase.java:111)
> at org.apache.jena.atlas.iterator.Iter$4.hasNext(Iter.java:264)
> at
> org.apache.jena.ext.com.google.common.collect.Iterators$5.hasNext(Iterators.java:547)
> at
> org.apache.jena.test.TransitiveEngineFailureTest$FailureTask.execQuery(TransitiveEngineFailureTest.java:109)
> at
> org.apache.jena.test.TransitiveEngineFailureTest$FailureTask.call(TransitiveEngineFailureTest.java:75)
> at
> org.apache.jena.test.TransitiveEngineFailureTest$FailureTask.call(TransitiveEngineFailureTest.java:1)
> at java.util.concurrent.FutureTask.run(FutureTask.java:266)
> at
> java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1142)
> at
> java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:617)
> at java.lang.Thread.run(Thread.java:745)
> {noformat}
> This failure is due to the
> {{org.apache.jena.reasoner.transitiveReasoner.TransitiveEngine}} class
> declaring the {{subClassAliases}}
> member variable (of type HashSet<Node>) as static. The prepare method
> re-creates this shared variable whenever {{isPrepared}}
> is false:
> {noformat}
> private void prepare() {
> if (isPrepared) return;
> subClassAliases = new HashSet<>();
> . . .
> isPrepared = true;
> }
> {noformat}
> In particular, the static {{subClassAliases}} variable will be re-created
> whenever a {{TransitiveEngine}} object is created
> ({{isPrepared}} is false at creation time). Note that HashSet is neither
> thread-safe nor being accessed in a thread-safe manner.
> The combination of attempting to remove triples from the model and attempting
> to execute queries on the model leads
> to unsafe accesses to {{subClassAliases}}, which eventually leads to a null
> pointer exception.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
