When looking around, TestStressIndexing has the same problem (only the time
is there 1.0 sec). Attached is a patch also for that. Also this test does
not use a volatile variable for the error boolean. Attached is a patch.

There are also other threads that may fail to do anything when the system is
under heavy load. Search for "currentTimeMillis" in tests :(

-----
Uwe Schindler
H.-H.-Meier-Allee 63, D-28213 Bremen
http://www.thetaphi.de
eMail: u...@thetaphi.de


> -----Original Message-----
> From: Uwe Schindler [mailto:u...@thetaphi.de]
> Sent: Friday, December 18, 2009 8:16 AM
> To: java-dev@lucene.apache.org
> Subject: RE: Build failed in Hudson: Lucene-trunk #1034
> 
> Here the patch, Mike, does this look ok?
> 
> When committed we can start a new Hudson run, if you like :-)
> 
> -----
> Uwe Schindler
> H.-H.-Meier-Allee 63, D-28213 Bremen
> http://www.thetaphi.de
> eMail: u...@thetaphi.de
> 
> > -----Original Message-----
> > From: Uwe Schindler [mailto:u...@thetaphi.de]
> > Sent: Friday, December 18, 2009 7:56 AM
> > To: java-dev@lucene.apache.org
> > Subject: RE: Build failed in Hudson: Lucene-trunk #1034
> >
> > > A new strange test failure:
> > >     [junit] Testcase:
> > > testDuringAddDelete(org.apache.lucene.index.TestIndexWriterReader):
> > > FAILED
> > >     [junit] null
> > >     [junit] junit.framework.AssertionFailedError: null
> > >     [junit]       at
> > >
> >
> org.apache.lucene.index.TestIndexWriterReader.testDuringAddDelete(TestInde
> > > xW
> > > riterReader.java:835)
> > >     [junit]       at
> > > org.apache.lucene.util.LuceneTestCase.runBare(LuceneTestCase.java:208)
> > >     [junit]
> > >     [junit]
> > >     [junit] Test org.apache.lucene.index.TestIndexWriterReader FAILED
> > >
> > >
> > > (occurred in test-tag!)
> > >
> > > It's again a multi-threaded test, after joining all threads, the sum
> of
> > > hits
> > > is not >0, so added documents were not seen.
> >
> > The problem is, the same like with the Benchmark test. The test spawns a
> > number of threads. All these threads run half a second and add Documents
> > during that time. The main thread does reopen and searches, and after
> 0.5
> > secs it joins all threads. It seems that because of clover analysis and
> > the
> > very slow machine (load!) on Hudson, all threads exited after starting
> > because creating the thread exceeded the limit on getCurrentTimeMillies.
> >
> > Two possibilities: Raise the wait time or better:
> > Replace by a do-while loop in the thread so each thread at least adds a
> > document one time.
> > In main thread also wait 0.5 secs, but after joining all threads again
> do
> > the search and add to sum.
> >
> > Must be done in trunk and tag.
> >
> > Uwe
> >
> >
> > ---------------------------------------------------------------------
> > To unsubscribe, e-mail: java-dev-unsubscr...@lucene.apache.org
> > For additional commands, e-mail: java-dev-h...@lucene.apache.org

Attachment: stresstest-timing.patch
Description: Binary data

---------------------------------------------------------------------
To unsubscribe, e-mail: java-dev-unsubscr...@lucene.apache.org
For additional commands, e-mail: java-dev-h...@lucene.apache.org

Reply via email to