jackrabbit-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Cech. Ulrich" <Ulrich.C...@aeb.de>
Subject AW: Indexer behaves differently while starting the repository?
Date Wed, 16 Mar 2011 13:55:24 GMT
I think I found something out:

The reason, because the indexer works while "storing" is, that there is actually only one
thread, which stores the data.
When the repository starts, the indexer starts multiple "ParsingTasks". I think, if many sessions
will store the data, there are also many indexer-threads, so the error would come up the same.

And the implementation of the AbstractStringBuffer.expandCapacity is very dangerous, because
if the char[] has 500MB, and we need 510, so the expandCapacity() method is called, then the
new char[] is initialized with 500MB * 2, which kills the error (or multiple ParsingTasks
expand from 100 to 200)

Can someone tell me, how you index very large files? Or is it a bad idea in general to fulltext-index
these big files?

Thanks in advance,
Ulrich


Mime
View raw message