[
https://issues.apache.org/jira/browse/LUCENE-1326?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12610315#action_12610315
]
Robert Newson commented on LUCENE-1326:
---------------------------------------
You're welcome and thanks for the quick response!
By the way, our specific issue was with DeflaterOutputStream.close(). It only
calls end() on the deflater if it created it itself, so this;
dos = new DeflaterOutputStream();
try {
...
} finally {
dos.close();
}
frees all resources but this does not;
dos = new DeflaterOutputStream(new Deflater(Deflater.BEST_COMPRESSION, true));
try {
...
} finally {
dos.close();
}
You have to call Deflater.end() *yourself* if you make one when you pass it in.
close() is not sufficient.
> Inflater.end() method not always called in FieldsReader
> -------------------------------------------------------
>
> Key: LUCENE-1326
> URL: https://issues.apache.org/jira/browse/LUCENE-1326
> Project: Lucene - Java
> Issue Type: Bug
> Affects Versions: 2.3.1
> Reporter: Robert Newson
>
> We've just found an insidious memory leak in our own application as we did
> not always call Deflater.end() and Inflater.end(). As documented here;
> http://bugs.sun.com/view_bug.do?bug_id=4797189
> The non-heap memory that the native zlib code uses is not freed in a timely
> manner.
> FieldsWriter appears safe as no exception can be thrown between the
> Deflater's creation and end() as it uses a ByteArrayOutputStream
> FieldsReader, however, is not safe. In the event of a DataFormatException the
> call to end() will not occur.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
