Mailing-List: contact java-user-help@lucene.apache.org; run by ezmlm
Precedence: bulk
Reply-To: java-user@lucene.apache.org
From: "Uwe Schindler" <uwe@thetaphi.de>
To: <java-user@lucene.apache.org>
References: <CADuAvzdx1D0VxjYtQgaVig=uqHGxjxxP491GvxgBRe7f-LEOng@mail.gmail.com>
In-Reply-To: <CADuAvzdx1D0VxjYtQgaVig=uqHGxjxxP491GvxgBRe7f-LEOng@mail.gmail.com>
Subject: RE: Weird corruption symptom, not making sense
Date: Mon, 16 Jan 2017 23:31:49 +0100
Message-ID: <014b01d27048$54624bd0$fd26e370$@thetaphi.de>
MIME-Version: 1.0
Content-Type: text/plain;
	charset="utf-8"
Content-Transfer-Encoding: quoted-printable
Content-Language: de
Thread-Index: AQIujHPerH9ISnJ1ZYY1LVvJ6osrU6CDd6Yg
archived-at: Mon, 16 Jan 2017 22:32:08 -0000

...or a JVM bug. We have seen those around PagedBytes in the past. What =
Java version?

-----
Uwe Schindler
Achterdiek 19, D-28357 Bremen
http://www.thetaphi.de
eMail: uwe@thetaphi.de

> -----Original Message-----
> From: Trejkaz [mailto:trejkaz@trypticon.org]
> Sent: Monday, January 16, 2017 11:27 PM
> To: Lucene Users Mailing List <java-user@lucene.apache.org>
> Subject: Weird corruption symptom, not making sense
>=20
> I have this thing where our UninvertingReader is getting an
> ArrayIndexOutOfBoundsException in production. I'm sure the index is
> corrupt, but I tried investigating the code and it still seems a bit
> odd.
>=20
> Caused by: java.lang.ArrayIndexOutOfBoundsException: -48116
> at org.apache.lucene.util.PagedBytes$Reader.fill(PagedBytes.java:118)
> at OurFieldCacheImpl.BinaryDocValuesImpl.get(SourceFile:844)
>=20
> In BinaryDocValuesImpl :
>=20
>     return new BinaryDocValues()
>     {
>         @Override
>         public BytesRef get(int docID)
>         {
>             int pointer =3D (int) docToOffset.get(docID);
>             if (pointer =3D=3D 0) {
>                 term.length =3D 0;
>             } else {
>                 bytes.fill(term, pointer);
>             }
>             return term;
>         }
>     };
>=20
> So "pointer" is the negative value presumably? Implying that somehow a
> negative value got into the docToOffset mappings.
> The value it's putting in comes from:
>=20
>     long pointer =3D bytes.copyUsingLengthPrefix(term);
>     postingsEnum =3D termsEnum.postings(postingsEnum,
> PostingsEnum.NONE);
>     while (true) {
>         int docID =3D postingsEnum.nextDoc();
>         if (docID =3D=3D DocIdSetIterator.NO_MORE_DOCS) {
>             break;
>         }
>         docToOffset.set(docID, pointer);
>     }
>=20
> So it seems like bytes.copyUsingLengthPrefix can return a negative
> value? But I looked in PagedBytes and couldn't see an obvious way to
> get a negative value.
>=20
> Is it possible some kind of overflow is happening here?
>=20
> TX
>=20
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: java-user-unsubscribe@lucene.apache.org
> For additional commands, e-mail: java-user-help@lucene.apache.org


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