Mailing-List: contact dev-help@lucene.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@lucene.apache.org
Received-SPF: neutral (athena.apache.org: local policy)
MIME-Version: 1.0
In-Reply-To: <00fb01ccc95e$bf153a50$3d3faef0$@thetaphi.de>
References: <2100754996.4751325391026356.JavaMail.hudson@aegis>
	<CA+A76TySdh_DSbstyc7DUf1x_KCVw2wqi4U6wRcyg_UH3M06ng@mail.gmail.com>
	<00fb01ccc95e$bf153a50$3d3faef0$@thetaphi.de>
Date: Mon, 2 Jan 2012 10:39:03 -0500
Message-ID: 
 <CAL8PwkbpjT1WbLYM=v0FOa+2yyXArYkWekYjAJsJjTtVkg-21w@mail.gmail.com>
Subject: Re: [JENKINS] Lucene-trunk - Build # 1786 - Failure
From: Michael McCandless <lucene@mikemccandless.com>
To: dev@lucene.apache.org
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable

+1 to using TotalHitCountCollector in tests that just need the hit
count, hmm though that might lower test coverage/exercise of the
normal collectors...

But in the mean time we should commit the quick fix so the test stops faili=
ng?

What changed, though?  Like why suddenly is this test failing so frequently=
...?

Mike McCandless

http://blog.mikemccandless.com

On Mon, Jan 2, 2012 at 9:56 AM, Uwe Schindler <uwe@thetaphi.de> wrote:
> Hi again,
>
>
>
> In my opinion, the test should simply test TopDocs.totalCount instead of =
the
> array size. There are more tests doing this. Maybe we should change all
> those tests that only count hits (and are not interested in the ScoreDocs=
 at
> all) to use TotalHitCountCollector instead of TopDocs?
>
>
>
> I would prefer the latter, I could open an issue.
>
>
>
> Uwe
>
> -----
>
> Uwe Schindler
>
> H.-H.-Meier-Allee 63, D-28213 Bremen
>
> http://www.thetaphi.de
>
> eMail: uwe@thetaphi.de
>
>
>
> From: martijn.is.hier@gmail.com [mailto:martijn.is.hier@gmail.com] On Beh=
alf
> Of Martijn v Groningen
> Sent: Sunday, January 01, 2012 12:50 PM
> To: dev@lucene.apache.org
> Subject: Re: [JENKINS] Lucene-trunk - Build # 1786 - Failure
>
>
>
> I think this is just a broken test case. If this test is ran with:
> "-Dtests.seed=3D-289aae8d40093437:-2c1c9ffc76ccb3bd:71f64018e9abbebb=A0-D=
tests.multiplier=3D3
> -Dtests.nightly=3Dtrue" then 1300 documents with term "aaa" are indexed.
> During searching the maximum number of documents to retrieve is hard code=
d
> to 1000.
>
> In that case the assertion on line 460 fails.
>
>
>
> Replacing:
>
> ScoreDoc[] hits =3D searcher.search(new TermQuery(new Term("field", "aaa"=
)),
> null, 1000).scoreDocs;
>
> with:
>
> ScoreDoc[] hits =3D searcher.search(new TermQuery(new Term("field", "aaa"=
)),
> null, n*100).scoreDocs;
>
> will fix this failure.
>
>
>
> Martijn
>
> On 1 January 2012 05:10, Apache Jenkins Server <jenkins@builds.apache.org=
>
> wrote:
>
> Build: https://builds.apache.org/job/Lucene-trunk/1786/
>
> 1 tests failed.
> REGRESSION: =A0org.apache.lucene.index.TestIndexWriter.testDiverseDocs
>
> Error Message:
> expected:<1300> but was:<1000>
>
> Stack Trace:
> junit.framework.AssertionFailedError: expected:<1300> but was:<1000>
> =A0 =A0 =A0 =A0at
> org.apache.lucene.util.LuceneTestCaseRunner.runChild(LuceneTestCaseRunner=
.java:165)
> =A0 =A0 =A0 =A0at
> org.apache.lucene.util.LuceneTestCaseRunner.runChild(LuceneTestCaseRunner=
.java:57)
> =A0 =A0 =A0 =A0at
> org.apache.lucene.index.TestIndexWriter.testDiverseDocs(TestIndexWriter.j=
ava:459)
> =A0 =A0 =A0 =A0at
> org.apache.lucene.util.LuceneTestCase$3$1.evaluate(LuceneTestCase.java:52=
8)
>
>
>
>
> Build Log (for compile errors):
> [...truncated 12691 lines...]
>
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscribe@lucene.apache.org
> For additional commands, e-mail: dev-help@lucene.apache.org

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