Hmmm. It is interesting, because that specific call (using
IndexWriter to "create" an index) was one of the causes in
LUCENE-140. But I'm pretty sure we fixed that cause. As part of
LUCENE-140 we also added further checks to catch re-using of an old
.del file at a lower level, and you're not hitting those.
After you close that IndexWriter, can you list the files in your
directory (that's a RAMDirectory right?)? Something like this:
import java.util.Arrays;
String[] l = directory.list();
Arrays.sort(l);
for(int i=0;i<l.length;i++)
System.out.println("file " + i + " = " + l[i] + " " +
directory.fileLength(l[i]) + " bytes");
Mike
"testn" <[EMAIL PROTECTED]> wrote:
>
> Does it help you to find out if I create an empty index before start the
> real
> operation?
>
> IndexWriter writer = new IndexWriter(directory, new
> SimpleAnalyzer(), true);
> writer.close();
> /* add new index afterward */
>
> This is to clean up the index since springmodules doesn't support the way
> to
> clear up the existing index.
>
>
>
> Michael McCandless-2 wrote:
> >
> >
> > OK. Is it possible to capture this as small test case?
> >
> > Maybe also call IndexWriter.setInfoStream(System.out) and capture details
> > on
> > what segments are being merged?
> >
> > Can you shed some light on how the application is using Lucene? Are you
> > doing
> > deletes as well as adds? Opening readers against this RAMDirectory?
> > Closing/
> > opening writers at different times? Any changes to the default parameters
> > (mergeFactor, maxBufferedDocs, etc.)?
> >
> > Mike
> >
> > "testn" <[EMAIL PROTECTED]> wrote:
> >>
> >> Here you go....
> >>
> >> -> Error during the indexing : docs out of order (0 <= 0 )
> >> org.apache.lucene.index.CorruptIndexException: docs out of order (0 <= 0
> >> )
> >> at
> >> org.apache.lucene.index.SegmentMerger.appendPostings(SegmentMerger.java:368)
> >> at
> >> org.apache.lucene.index.SegmentMerger.mergeTermInfo(SegmentMerger.java:325)
> >> at
> >> org.apache.lucene.index.SegmentMerger.mergeTermInfos(SegmentMerger.java:297)
> >> at
> >> org.apache.lucene.index.SegmentMerger.mergeTerms(SegmentMerger.java:261)
> >> at
> >> org.apache.lucene.index.SegmentMerger.merge(SegmentMerger.java:98)
> >> at
> >> org.apache.lucene.index.IndexWriter.mergeSegments(IndexWriter.java:1883)
> >> at
> >> org.apache.lucene.index.IndexWriter.flushRamSegments(IndexWriter.java:1741)
> >> at
> >> org.apache.lucene.index.IndexWriter.flushRamSegments(IndexWriter.java:1733)
> >> at
> >> org.apache.lucene.index.IndexWriter.maybeFlushRamSegments(IndexWriter.java:1727)
> >> at
> >> org.apache.lucene.index.IndexWriter.addDocument(IndexWriter.java:1004)
> >> at
> >> org.apache.lucene.index.IndexWriter.addDocument(IndexWriter.java:983)
> >> at
> >> org.springmodules.lucene.index.factory.SimpleLuceneIndexWriter.addDocument(SimpleLuceneIndexWriter.java:44)
> >> )
> >> at
> >> org.springmodules.lucene.index.object.database.DefaultDatabaseIndexer.doHandleRequest(DefaultDatabaseIndexer.java:306)
> >> at
> >> org.springmodules.lucene.index.object.database.DefaultDatabaseIndexer.index(DefaultDatabaseIndexer.java:354)
> >>
> >>
> >> Michael McCandless-2 wrote:
> >> >
> >> >
> >> > Well then that is particularly spooky!!
> >> >
> >> > And, hopefully, possible/easy to reproduce. Thanks.
> >> >
> >> > Mike
> >> >
> >> > "testn" <[EMAIL PROTECTED]> wrote:
> >> >>
> >> >> I use RAMDirectory and the error often shows the low number. Last time
> >> it
> >> >> happened with message "7<=7". Nest time it happens, I will try to
> >> capture
> >> >> the stacktrace.
> >> >>
> >> >>
> >> >>
> >> >> Michael McCandless-2 wrote:
> >> >> >
> >> >> >
> >> >> > "testn" <[EMAIL PROTECTED]> wrote:
> >> >> >>
> >> >> >> Using Lucene 2.2.0, I still sporadically got doc out of order
> >> error. I
> >> >> >> indexed all of my stuff in one thread. Do you have any idea why it
> >> >> >> happens?
> >> >> >
> >> >> > Hmmmmm, that is not good. I thought we had finally fixed this with
> >> >> > LUCENE-140. Though un-corrected disk errors could in theory lead to
> >> >> > this too.
> >> >> >
> >> >> > Are you able to easily reproduce it? Can you post the full
> >> exception?
> >> >> >
> >> >> > Mike
> >> >> >
> >> >> >
> >> ---------------------------------------------------------------------
> >> >> > To unsubscribe, e-mail: [EMAIL PROTECTED]
> >> >> > For additional commands, e-mail: [EMAIL PROTECTED]
> >> >> >
> >> >> >
> >> >> >
> >> >>
> >> >> --
> >> >> View this message in context:
> >> >> http://www.nabble.com/out-of-order-tf4276385.html#a12173705
> >> >> Sent from the Lucene - Java Users mailing list archive at Nabble.com.
> >> >>
> >> >>
> >> >> ---------------------------------------------------------------------
> >> >> To unsubscribe, e-mail: [EMAIL PROTECTED]
> >> >> For additional commands, e-mail: [EMAIL PROTECTED]
> >> >>
> >> >
> >> > ---------------------------------------------------------------------
> >> > To unsubscribe, e-mail: [EMAIL PROTECTED]
> >> > For additional commands, e-mail: [EMAIL PROTECTED]
> >> >
> >> >
> >> >
> >>
> >> --
> >> View this message in context:
> >> http://www.nabble.com/out-of-order-tf4276385.html#a12184067
> >> Sent from the Lucene - Java Users mailing list archive at Nabble.com.
> >>
> >>
> >> ---------------------------------------------------------------------
> >> To unsubscribe, e-mail: [EMAIL PROTECTED]
> >> For additional commands, e-mail: [EMAIL PROTECTED]
> >>
> >
> > ---------------------------------------------------------------------
> > To unsubscribe, e-mail: [EMAIL PROTECTED]
> > For additional commands, e-mail: [EMAIL PROTECTED]
> >
> >
> >
>
> --
> View this message in context:
> http://www.nabble.com/out-of-order-tf4276385.html#a12186579
> Sent from the Lucene - Java Users mailing list archive at Nabble.com.
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [EMAIL PROTECTED]
> For additional commands, e-mail: [EMAIL PROTECTED]
>
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
