lucene-java-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "yuexiang zhang" <easyj...@gmail.com>
Subject java.lang.OutOfMemoryError when search in large index
Date Sat, 08 Jul 2006 07:50:30 GMT
I have finished importing 10,000,000 docments into Lucene 2.0 .

Every  document includes  journalTitle and articleTitle both of which should
be sortable, so I
add other two fields : journalTitle_sort and  articleTitle_sort both of
which use un_tokenized option for
indexer.

When I   query by "journaltitle:advertising" and sort by journaltitle_sort,
it 's ok for return "total result num is 28684"
But when I use the same query and change sort field from journaltitle_sort
to articleTitle_sort, java.lang.OutOfMemoryError  happens.
Then I try set JVM options to "-Xmx2048m", it also do take affect,
java.lang.OutOfMemoryError  happens too.

I think maybe lucene preload all  unique sortable field vlaue into memory,
but my articleTitle_sort has too large different values because of
un_tokenized  option.

Who can help me? Thanks!

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message