[
https://issues.apache.org/jira/browse/LUCENE-4953?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13640979#comment-13640979
]
Uwe Schindler edited comment on LUCENE-4953 at 4/24/13 9:29 PM:
----------------------------------------------------------------
bq. Ie, today LuceneTestCase.maybeWrapReader never incRefs the incoming reader,
but with the fix, ParallelCompositeReader now does ... so this leads to
failures.
Passing false instead of true to the wrapper is wrong. The solution is to leave
the construction as is, just in doClose() call notifyReaderClosedListeners() on
all subreaders (which is unfortunately private). But this would still not help
for tests, because maybeWrapReader tries to hide itsself from FieldCache (this
is why it has FCInvisibleMultiReader). Tests almost always close the original
reader and never the wrapper, so with this patch the whole thing does not work.
The test failure is very seldom because it only happens:
- if you wrap (rarely) with a ParallelMultiReader (more rarely)
- use FieldCache
The number of tests with FieldCache is very small, so it took more than 1 year
to see the first failure :-)
In fact the problem for the actual test failure is maybeWrapReader in
combination with FieldCache. maybeWrapReader should not use
ParallelCompositeReader, if it knows that FieldCache will be used.
Unfortunately this is not known before.
The problem with the readerClosedListener not called by ParallelCompositeReader
(with closeSubReaders=true) is real, and only affects PCR, because it creates
atomic readers with "own" fieldcache key. Because of that it should manually
unregister them (and not close them)
was (Author: thetaphi):
bq. Ie, today LuceneTestCase.maybeWrapReader never incRefs the incoming
reader, but with the fix, ParallelCompositeReader now does ... so this leads to
failures.
Passing false instead of true to the wrapper is wrong. The solution is to leave
the construction as is, just in doClose() call notifyReaderClosedListeners() on
all subreaders (which is unfortunately private). But this would still not help
for tests, because maybeWrapReader tries to hide itsself from FieldCache (this
is why it has FCInvisibleMultiReader). Tests always class the original reader
never the wrapper, so although with this patch the whole thing does not work.
The test failure is very seldom because it only happens:
- if you wrap (rarely) with a ParallelMultiReader (more rarely)
- use FieldCache
The number of tests with FieldCache is very small, so it took more than 1 year
to see the first failure :-)
In fact the problem for the actual test failure is maybeWrapReader in
combination with FieldCache. maybeWrapReader should not use
ParallelCompositeReader, if it knows that FieldCache will be used.
Unfortunately this is not known before.
The problem with the readerClosedListener not called by ParallelCompositeReader
(with closeSubReaders=true) is real, and only affects PCR, because it creates
atomic readers with "own" fieldcache key. Because of that it should manually
unregister them (and not close them)
> readerClosedListener is not invoked for ParallelCompositeReader's leaves
> ------------------------------------------------------------------------
>
> Key: LUCENE-4953
> URL: https://issues.apache.org/jira/browse/LUCENE-4953
> Project: Lucene - Core
> Issue Type: Bug
> Reporter: Michael McCandless
> Fix For: 5.0, 4.4
>
> Attachments: LUCENE-4953.patch, LUCENE-4953.patch
>
>
> There was a test failure last night:
> {noformat}
> 1 tests failed.
> REGRESSION:
> org.apache.lucene.search.grouping.AllGroupHeadsCollectorTest.testBasic
> Error Message:
> testBasic(org.apache.lucene.search.grouping.AllGroupHeadsCollectorTest):
> Insane FieldCache usage(s) found expected:<0> but was:<2>
> Stack Trace:
> java.lang.AssertionError:
> testBasic(org.apache.lucene.search.grouping.AllGroupHeadsCollectorTest):
> Insane FieldCache usage(s) found expected:<0> but was:<2>
> at
> __randomizedtesting.SeedInfo.seed([1F9C2A2AD23A8E02:B466373F0DE6082C]:0)
> at org.junit.Assert.fail(Assert.java:93)
> at org.junit.Assert.failNotEquals(Assert.java:647)
> at org.junit.Assert.assertEquals(Assert.java:128)
> at org.junit.Assert.assertEquals(Assert.java:472)
> at
> org.apache.lucene.util.LuceneTestCase.assertSaneFieldCaches(LuceneTestCase.java:592)
> at
> org.apache.lucene.util.TestRuleFieldCacheSanity$1.evaluate(TestRuleFieldCacheSanity.java:55)
> at
> org.apache.lucene.util.AbstractBeforeAfterRule$1.evaluate(AbstractBeforeAfterRule.java:46)
> at
> com.carrotsearch.randomizedtesting.rules.SystemPropertiesInvariantRule$1.evaluate(SystemPropertiesInvariantRule.java:55)
> at
> org.apache.lucene.util.TestRuleThreadAndTestName$1.evaluate(TestRuleThreadAndTestName.java:49)
> at
> org.apache.lucene.util.TestRuleIgnoreAfterMaxFailures$1.evaluate(TestRuleIgnoreAfterMaxFailures.java:70)
> at
> org.apache.lucene.util.TestRuleMarkFailure$1.evaluate(TestRuleMarkFailure.java:48)
> at
> com.carrotsearch.randomizedtesting.rules.StatementAdapter.evaluate(StatementAdapter.java:36)
> at
> com.carrotsearch.randomizedtesting.ThreadLeakControl$StatementRunner.run(ThreadLeakControl.java:358)
> at
> com.carrotsearch.randomizedtesting.ThreadLeakControl.forkTimeoutingTask(ThreadLeakControl.java:782)
> at
> com.carrotsearch.randomizedtesting.ThreadLeakControl$3.evaluate(ThreadLeakControl.java:442)
> at
> com.carrotsearch.randomizedtesting.RandomizedRunner.runSingleTest(RandomizedRunner.java:746)
> at
> com.carrotsearch.randomizedtesting.RandomizedRunner$3.evaluate(RandomizedRunner.java:648)
> at
> com.carrotsearch.randomizedtesting.RandomizedRunner$4.evaluate(RandomizedRunner.java:682)
> at
> com.carrotsearch.randomizedtesting.RandomizedRunner$5.evaluate(RandomizedRunner.java:693)
> at
> org.apache.lucene.util.AbstractBeforeAfterRule$1.evaluate(AbstractBeforeAfterRule.java:46)
> at
> org.apache.lucene.util.TestRuleStoreClassName$1.evaluate(TestRuleStoreClassName.java:42)
> at
> com.carrotsearch.randomizedtesting.rules.SystemPropertiesInvariantRule$1.evaluate(SystemPropertiesInvariantRule.java:55)
> at
> com.carrotsearch.randomizedtesting.rules.NoShadowingOrOverridesOnMethodsRule$1.evaluate(NoShadowingOrOverridesOnMethodsRule.java:39)
> at
> com.carrotsearch.randomizedtesting.rules.NoShadowingOrOverridesOnMethodsRule$1.evaluate(NoShadowingOrOverridesOnMethodsRule.java:39)
> at
> com.carrotsearch.randomizedtesting.rules.StatementAdapter.evaluate(StatementAdapter.java:36)
> at
> org.apache.lucene.util.TestRuleAssertionsRequired$1.evaluate(TestRuleAssertionsRequired.java:43)
> at
> org.apache.lucene.util.TestRuleMarkFailure$1.evaluate(TestRuleMarkFailure.java:48)
> at
> org.apache.lucene.util.TestRuleIgnoreAfterMaxFailures$1.evaluate(TestRuleIgnoreAfterMaxFailures.java:70)
> at
> org.apache.lucene.util.TestRuleIgnoreTestSuites$1.evaluate(TestRuleIgnoreTestSuites.java:55)
> at
> com.carrotsearch.randomizedtesting.rules.StatementAdapter.evaluate(StatementAdapter.java:36)
> at
> com.carrotsearch.randomizedtesting.ThreadLeakControl$StatementRunner.run(ThreadLeakControl.java:358)
> at java.lang.Thread.run(Thread.java:722)
> Build Log:
> [...truncated 6904 lines...]
> [junit4:junit4] Suite:
> org.apache.lucene.search.grouping.AllGroupHeadsCollectorTest
> [junit4:junit4] 2> *** BEGIN
> testBasic(org.apache.lucene.search.grouping.AllGroupHeadsCollectorTest):
> Insane FieldCache usage(s) ***
> [junit4:junit4] 2> VALUEMISMATCH: Multiple distinct value objects for
> ParallelAtomicReader(_0(5.0):C3)+id
> [junit4:junit4] 2> 'ParallelAtomicReader(_0(5.0):C3)'=>'id',class
> org.apache.lucene.index.SortedDocValues,0.5=>org.apache.lucene.search.FieldCacheImpl$SortedDocValuesImpl#386041791
> (size =~ 232 bytes)
> [junit4:junit4] 2>
> 'ParallelAtomicReader(_0(5.0):C3)'=>'id',int,org.apache.lucene.search.FieldCache.DEFAULT_INT_PARSER=>org.apache.lucene.search.FieldCacheImpl$IntsFromArray#140912913
> (size =~ 48 bytes)
> [junit4:junit4] 2>
> 'ParallelAtomicReader(_0(5.0):C3)'=>'id',int,null=>org.apache.lucene.search.FieldCacheImpl$IntsFromArray#140912913
> (size =~ 48 bytes)
> [junit4:junit4] 2>
> [junit4:junit4] 2> VALUEMISMATCH: Multiple distinct value objects for
> ParallelAtomicReader(_1(5.0):C5)+id
> [junit4:junit4] 2>
> 'ParallelAtomicReader(_1(5.0):C5)'=>'id',int,null=>org.apache.lucene.search.FieldCacheImpl$IntsFromArray#1105632232
> (size =~ 56 bytes)
> [junit4:junit4] 2>
> 'ParallelAtomicReader(_1(5.0):C5)'=>'id',int,org.apache.lucene.search.FieldCache.DEFAULT_INT_PARSER=>org.apache.lucene.search.FieldCacheImpl$IntsFromArray#1105632232
> (size =~ 56 bytes)
> [junit4:junit4] 2> 'ParallelAtomicReader(_1(5.0):C5)'=>'id',class
> org.apache.lucene.index.SortedDocValues,0.5=>org.apache.lucene.search.FieldCacheImpl$SortedDocValuesImpl#27148040
> (size =~ 232 bytes)
> [junit4:junit4] 2>
> [junit4:junit4] 2> *** END
> testBasic(org.apache.lucene.search.grouping.AllGroupHeadsCollectorTest):
> Insane FieldCache usage(s) ***
> [junit4:junit4] 2> NOTE: download the large Jenkins line-docs file by
> running 'ant get-jenkins-line-docs' in the lucene directory.
> [junit4:junit4] 2> NOTE: reproduce with: ant test
> -Dtestcase=AllGroupHeadsCollectorTest -Dtests.method=testBasic
> -Dtests.seed=1F9C2A2AD23A8E02 -Dtests.multiplier=2 -Dtests.nightly=true
> -Dtests.slow=true
> -Dtests.linedocsfile=/home/hudson/lucene-data/enwiki.random.lines.txt
> -Dtests.locale=be_BY -Dtests.timezone=Asia/Manila
> -Dtests.file.encoding=US-ASCII
> [junit4:junit4] FAILURE 0.75s J1 | AllGroupHeadsCollectorTest.testBasic <<<
> [junit4:junit4] > Throwable #1: java.lang.AssertionError:
> testBasic(org.apache.lucene.search.grouping.AllGroupHeadsCollectorTest):
> Insane FieldCache usage(s) found expected:<0> but was:<2>
> [junit4:junit4] > at
> __randomizedtesting.SeedInfo.seed([1F9C2A2AD23A8E02:B466373F0DE6082C]:0)
> [junit4:junit4] > at
> org.apache.lucene.util.LuceneTestCase.assertSaneFieldCaches(LuceneTestCase.java:592)
> [junit4:junit4] > at java.lang.Thread.run(Thread.java:722)
> [junit4:junit4] 2> NOTE: test params are: codec=Lucene42:
> {sort3=MockFixedIntBlock(blockSize=733), id=Pulsing41(freqCutoff=3
> minBlockSize=50 maxBlockSize=177), content=MockFixedIntBlock(blockSize=733),
> author=Pulsing41(freqCutoff=3 minBlockSize=50 maxBlockSize=177),
> sort2=MockVariableIntBlock(baseBlockSize=71), sort1=Pulsing41(freqCutoff=3
> minBlockSize=50 maxBlockSize=177), group=Pulsing41(freqCutoff=3
> minBlockSize=50 maxBlockSize=177)},
> docValues:{author_dv=DocValuesFormat(name=Disk),
> group_dv=DocValuesFormat(name=Disk)},
> sim=RandomSimilarityProvider(queryNorm=false,coord=yes): {content=IB LL-L1,
> author=DFR GBZ(0.3)}, locale=be_BY, timezone=Asia/Manila
> [junit4:junit4] 2> NOTE: FreeBSD 9.0-RELEASE amd64/Oracle Corporation
> 1.7.0_17 (64-bit)/cpus=16,threads=1,free=157973280,total=249626624
> [junit4:junit4] 2> NOTE: All tests run in this JVM:
> [GroupFacetCollectorTest, AllGroupsCollectorTest, AllGroupHeadsCollectorTest]
> {noformat}
> It reproduces, and happens because ParallelCompositeReader isn't invoking the
> reader listeners on its .leaves() when everything is closed. I made a
> separate test case to show the issue ...
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org
For additional commands, e-mail: dev-h...@lucene.apache.org
