IBM CICS Performance Series: Comparing Type 2 and Type 4 JDBC Driver Performance with IBM CICS Transaction Server for z/OS V5.2 Liberty JVM server

An IBM Redpaper publication

thumbnail 

Published on August 28, 2015

  1. .PDF (1.4 MB)

Share this page:   

ISBN-10: 0738454389
ISBN-13: 9780738454382
IBM Form #: REDP-5208-00


Authors: Graham Rawson

    menu icon

    Abstract

    This IBM® Redpaper™ publication is a study of the performance of a Java Platform, Enterprise Edition (Java EE) application that is hosted by an IBM CICS® Transaction Server for z/OS® (CICS) V5.2 Liberty Java virtual machine (JVM) server that connects to an IBM DB2® for z/OS database server. The use of type 2 and type 4 Java Database Connectivity (JDBC) drivers supplied by DB2 for z/OS is described, and performance comparisons are made using processor (CPU) resource use and response time data.

    CPU usage data includes measurement of offload eligibility using IBM System z® Integrated Information Processors (zIIPs).

    The target audience for this IBM Redpaper publication is as follows:

    • Systems Architects considering the performance characteristics of using either the type 2 or type 4 JDBC driver in a CICS Liberty environment.
    • Capacity Planners and Performance Analysts trying to understand how to configure the JDBC drivers and measure the resource usage and response time characteristics of the JDBC drivers.

    This paper contains a description of the Java EE application used in this performance study, and a description of the environment on which the performance measurements were made. The results from the performance measurements and an analysis of these results are presented.

    This paper covers the following topics:

    • A description of the Java EE application used, and a description of some minor changes made to extend the application functionality
    • A description of the hardware and software comprising the measurement environment
    • An overview of the JDBC driver types used in this study
    • A description of database connection sharing performance optimizations used in this study
    • An explanation of the configuration of a JDBC driver type for use in a CICS Liberty JVM server
    • An explanation of the performance metrics used in this study
    • Presentation and analysis of the results in terms of CPU cost, zIIP eligibility, and request response time

    This paper is one in a series focused on CICS performance. It is written by members of the IBM Hursley CICS development community. The subject matter in this series is based on feedback from CICS customers.

    Table of Contents

    Introduction

    Executive summary

    The TradeLite application

    The test environment

    JDBC driver types

    Configuring a JDBC driver type for a CICS Liberty JVM server

    JDBC driver connection reuse mechanisms

    Performance measurement procedure

    Configurations measured

    Transaction scope in the TradeLite application

    Results

    Conclusion

     

    Others who read this also read