jackrabbit-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Sandro Boehme <sandro.boe...@gmx.de>
Subject Re: workspace configuration for adding 1.000.000 nodes
Date Tue, 21 Oct 2008 11:00:21 GMT
Hello again,

just not to cause uncertainness: Jackrabbit runs stable.
In the meantime I tested my use case creating 1.000.000 nodes with the 
configuration attached and with no dependency to my application code. 
The result was that Jackrabbit did not get an OutOfMemoryError.



Sandro Boehme schrieb:
> Hello,
> for a proof of concept I try to create 1.000.000 nodes with not more 
> than 300 child nodes under one node. Session.save() is called after 
> every 100 nodes. But after almost 500.000 nodes I get an 
> "OutOfMemoryError: Java heap space" within a Lucene method (at 
> org.apache.lucene.index.SegmentReader.createFakeNorms(SegmentReader.java:426)). 
> I use RMI to connect to a usual Jackrabbit 1.4 Tomcat installation. The 
> server is configured to start with export 
> JAVA_OPTS="-XX:MaxPermSize=256m -Xmx1024m" and I use the attached 
> configuration files.
> I guess there is something wrong with my <SearchIndex/> configuration in 
> the workspace.xml file. But I could not find out what the problem is.
> If somebody has a configuration that should be able to handle the 
> creation of 1.000.000 nodes I would be glad if he could send it as a 
> reply. Thanks in advance,
> Sandro

View raw message