on 02-14-2005 4:07 PM
Hi All,
I've recently been on the SAP XI 3.0 Bootcamp in the UK and have an issue someone might be able to clear up. We were told that mapping is the biggest over head which makes sense. We were also told that complex mapping scenarios were more efficiently handled using XSLT mapping rather than Java Mapping via Message Mapping in the IR.
Has anyone got an opinion on this ?
Thanks
Gary
Thanks to all of you. The message is not particularly clear but I deduce from this that the consensus is that graphical/Java mapping is more efficient for complex mapping scenarios.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
When I say Java I am referring to the graphical mapping tool which I believe compiles into Java. Does this change things ?
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Hi
I agree with NW to a point.
Take in mind.
XSLT loads all the XML message into memory and then maps the data according to the instructions. This means that the process will use more memory.
The SAP XI message mapping only loads the node that it is working on and maps according to the instructions. This means it uses less memory.
But still depending on the complexity.
Cheers
Ian
Hi,
Java vs XSLT mapping is not an easy question to solve but anyway mapping performance should be better than XSLT for mapping with calculation.
Keep in mind that XSLT parsing is using character based processing so to do calculation with XSLT , the parser has to first convert all data into numeric which is not the case in Java using numeric as well as character data.
But anyway I agree that depends on the mapping complexity.
Greetings
Nicolas
Hi Gary,
Mapping should be done outside BPM, XSLT is faster than JAVA.
Cheers,
Paul
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
That depends really!!!
We do not have access to published benchmarks from SAP on the mapping performance of the XI Mapping Runtime, versus other XSLT implementations.
Hence, I will dispute if complex mapping should be done outside XI; there is no published metrix that compares XSLT performance with SAP.
Again, Java based mapping is faster than XSLT. In fact, we had to resort to custom mapping in Java when the xalan / xerces performance was unacceptably slow on a 25MB document for a certain scenario. This was of course using a different EAI product.
User | Count |
---|---|
95 | |
11 | |
10 | |
9 | |
9 | |
7 | |
6 | |
5 | |
5 | |
4 |
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.