Use of XSLT when dealing with Large XML

Hi All

I wanted to know whether I should use xslt or not when I am dealing with large xml.

IS is receiving large xml (it could be till 100 MB). Initially I used xml node iterator approach so less memory should be utilized.But node iterator approach was taking more processing time to perform mapping transformation.

So I used XSLT. XSLT is taking less processing time. But I am not sure whether XSLT is consuming more memory or not ? I wanted to know so I can decide that should I use XSLT while dealing with large xml and where transformation should be performed quickly OR is there any way to know it OR should I really monitor the memory consumption in IS ? Please suggest.


Use JVisualVM to monitor the memory/thread usages while you run the XSLT transformation services. During this test generate the thread dump via IS or from the Linux box and analyze.

If you are just mapping it to another format, xslt may be faster and using less memory.
But if you are doing other things, like saving to DB, sending to other systems, then you still need to do those in IS. keeping the xml as bytes or string will for sure leave less memory footprint, comparing to a parsed doc.