SOAP Message Size Performance Considerations

An IBM Redpaper publication

Published 29 August 2007

cover image

IBM Form #: REDP-4344-00
(62 pages)

More options

Rate and comment

Authors: Chris Rayns, Trevor Clarke, Mogens Conrad, Charlie Wiese


This IBM Redpaper publication examines performance when using different SOAP message sizes in CICS Web Services.
We used a preexisting basic Web Services environment with no message handlers or header handlers. Using WebSphere Studio Workload Simulator (WSWS), we ran a number of simulations using predefined scripts. For more information about WSWS, refer to WebSphere Studio Workload Simulator User’s Guide, SC31-6307. We then processed the CICS SMF data using CICS Performance Analyzer (CICS PA) and collated the statistics. For more information about CICSPA, see CICS Performance Analyzer User’s Guide, SC34-6307.
This paper also reviews a customer scenario, DanskeBank. We provide an overview of the bank’s ATM system, and briefly discuss the IFX framework. We detail our use of WebSphere Developer for zSeries (WD/z) to simulate request messages from an ATM, present a performance comparison between optimized and non-optimized schema versions, and cover schema versus container definitions, large fields, and schema complexity. Finally, we provide an example of the impact of mapping levels.

Table of contents

Chapter 1. SOAP message size performance considerations
Chapter 2. DanskeBank ATM application

Follow IBM Redbooks

Follow IBM Redbooks