You are viewing a plain text version of this content. The canonical link for it is here.
Posted to java-user@lucene.apache.org by Dan Gould <dl...@cs.brown.edu> on 2005/12/09 02:49:11 UTC
JVM Crash in Lucene
Hi--
I'm relatively new to Lucene. When I run my app, I get a JVM error.
This gets called a lot, but only fails every once in awhile (maybe 1 in
100 calls?)
I filed a report with Sun, but I don't expect to hear anything from them.
So, I was wondering if any Lucene experts have run across this or have any
ideas for workarounds (if not, I'll have to take lucene out of my system
and manage the termvectors myself):
#
# An unexpected error has been detected by HotSpot Virtual Machine:
#
# SIGSEGV (0xb) at pc=0x0000002a95a5c300, pid=26545, tid=1085655392
#
# Java VM: Java HotSpot(TM) 64-Bit Server VM (1.5.0_06-b05 mixed mode)
# Problematic frame:
# V [libjvm.so+0x4fa300]
#
# An error report file with more information is saved as
hs_err_pid26545.log
That error file is huge, but the salient part seems to be:
Stack: [0x0000000040a5c000,0x0000000040b5d000), sp=0x0000000040b59f70,
free sp
ace=1015k
Native frames: (J=compiled Java code, j=interpreted, Vv=VM code, C=native
code)
V [libjvm.so+0x4fa300]
V [libjvm.so+0x4fa3c2]
V [libjvm.so+0x4fa876]
V [libjvm.so+0x58051e]
V [libjvm.so+0x58017c]
V [libjvm.so+0x2b968b]
V [libjvm.so+0x2b67ba]
V [libjvm.so+0x25f3c7]
V [libjvm.so+0x2bed06]
V [libjvm.so+0x2be6b2]
V [libjvm.so+0x6163f9]
V [libjvm.so+0x610ffe]
V [libjvm.so+0x554745]
Current CompileTask:
opto:1621
org.apache.lucene.index.IndexReader$1.doBody()Ljava/lang/Object;
(99 bytes)
My code that is causing the JVM crash is:
//Get the terms and frequencies
IndexReader indexreader = null;
TermFreqVector tv = null;
try {
indexreader = indexManager.getIndexReader();
tv = indexManager.getTermVector(indexreader, item);
if(null==tv) {
if(log.isErrorEnabled()) log.error("No term vector for item " + item.getId());
//??? better handling
return;
}
}
finally {
//XXX should close earlier
if(null!=indexreader) {
try {
indexreader.close();
}
catch(IOException ex) {
if(log.isDebugEnabled()) log.debug("IOException cleaning up indexreader ", ex);
}
}
}
//Replace the String[] with its equivalent long[]... (and tally up total occurs)
String[] termstrings = tv.getTerms();
int[] freqs = tv.getTermFrequencies();
Thank you for your help,
Dan
---------------------------------------------------------------------
To unsubscribe, e-mail: java-user-unsubscribe@lucene.apache.org
For additional commands, e-mail: java-user-help@lucene.apache.org
Re: JVM Crash in Lucene
Posted by Yonik Seeley <ys...@gmail.com>.
The only problems I've had with 1.5 JVM crashes and Lucene was related
to stack overflow... try increasing the stack size and see of anything
different happens.
My crashes happened while trying to use Luke to open a 4GB index with
thousands of indexed fields.
-Yonik
---------------------------------------------------------------------
To unsubscribe, e-mail: java-user-unsubscribe@lucene.apache.org
For additional commands, e-mail: java-user-help@lucene.apache.org
Re: JVM Crash in Lucene
Posted by Chris Hostetter <ho...@fucit.org>.
: I'm relatively new to Lucene. When I run my app, I get a JVM error.
: This gets called a lot, but only fails every once in awhile (maybe 1 in
: 100 calls?)
i'm not that familiar with TermFreqVectors, and I have no idea what
indexManager is, but I'm suprised this works at all ... I thought calling
the getters on a TermFreqVector required it to read from the underlying
files (which would require that the reader still be open).
Is it possible that that's your problem? .. is there perhaps some strange
GC interaction issue that results in the data TermFreqVector is looking
for to go bye-bye because you closed the reader?
(it sounds silly when I read it back to myself ... but it's the best guess
i've got just reading your code)
: My code that is causing the JVM crash is:
:
: //Get the terms and frequencies
: IndexReader indexreader = null;
: TermFreqVector tv = null;
: try {
: indexreader = indexManager.getIndexReader();
: tv = indexManager.getTermVector(indexreader, item);
: if(null==tv) {
: if(log.isErrorEnabled()) log.error("No term vector for item " + item.getId());
: //??? better handling
: return;
: }
: }
: finally {
: //XXX should close earlier
: if(null!=indexreader) {
: try {
: indexreader.close();
: }
: catch(IOException ex) {
: if(log.isDebugEnabled()) log.debug("IOException cleaning up indexreader ", ex);
: }
: }
: }
:
: //Replace the String[] with its equivalent long[]... (and tally up total occurs)
: String[] termstrings = tv.getTerms();
: int[] freqs = tv.getTermFrequencies();
-Hoss
---------------------------------------------------------------------
To unsubscribe, e-mail: java-user-unsubscribe@lucene.apache.org
For additional commands, e-mail: java-user-help@lucene.apache.org