Skip to main content

Software > Networking >

Application Workload Modeler

Technical detail

IBM Application Workload Modeler for z/OS Release 1 provides the ability to model, measure, and analyze the performance of networks and applications in a client/server, multiprotocol, multiplatform environment. With Application Workload Modeler R1, you can more accurately plan for the roll-out of additional software or function, and determine where upgrades may be required in your network and systems.

Application Workload Modeler R1 can be used in two modes:

  • Client/Server mode benchmarks allow you to evaluate the performance of your communications stack (such as TCP/IP and SNA) and your network by modeling the behavior of an application and generating the associated network traffic.

    In this case, the actual application need not be installed. Application Workload Modeler can model both the client and server.

  • Application Client mode benchmarks allow you to measure the overall end-to-end performance of a well-known TCP/IP application server. This is done by using Application Workload Modeler as a client to an existing application server, such as a TN3270 or FTP server. Application Workload Modeler will generate real traffic, as specified by you, to the application server.

Application Workload Modeler Client/Server mode benchmarks support TCP/IP applications and networks for the current version of the Internet Protocol (IPv4) that is widely deployed. They also support the emerging next generation Internet Protocol (IPv6). This allows users to test and evaluate the performance of their IPv6 network infrastructure prior to the introduction of production level, IPv6 network traffic and applications.

In both modes of operation, Application Workload Modeler allows you to simulate large numbers of clients and stress workload levels from a single host. When even larger workload levels are required for a test, several hosts with Application Workload Modeler installed can be used to simulate the necessary traffic. In this scenario, a single Application Workload Modeler instance can act as the controller, providing a single control point that can be used to coordinate the execution of the test and the gathering of the results for all client and server Application Workload Modeler instances involved in the test.

The ability to simulate client activity concurrently, from several Application Workload Modeler instances, is also useful when attempting to model client application performance across different areas of the network. You can more accurately simulate client network activity by installing Application Workload Modeler instances on one or more hosts that are in close proximity, from a network perspective, to the representative client hosts that will be generating the application network traffic. This allows users to identify specific network bottlenecks that may affect only a subset of clients.

Application Workload Modeler provides additional flexibility in this area with its multiplatform support: z/OS, OS/390, Linux on zSeries, and Linux. Refer to the Software Requirements for the specific versions and releases supported.

Workload simulation tests can involve Application Workload Modeler on a single platform or Application Workload Modeler on any combination of the platforms supported. This allows you the flexibility of selecting the platform(s) based on the objectives of the specific workload modeling exercise. For example, if a user needs to model an application where the server resides on z/OS and the clients reside at the edges of the enterprise network, an Application Workload Modeler client could be run on a Linux host in close proximity to the client network and the application server could be simulated on a z/OS system within the enterprise network.

Services in support of Application Workload Modeler R1 are available from IBM Global Services and AIM zSeries Software Services.

The following offerings are available from AIM zSeries Software Services:

  • Installation, Configuration, Training, and Mentoring, which includes:
    • Helping you install and configure Application Workload Modeler
    • Two days of on-site support for customers using Application Workload Modeler in their own environment
    • Determining how to use the product most effectively for the design, performance, or capacity planning issues you seek to resolve
  • A Try and Buy Offering, which will allow you to engage the services team to use Application Workload Modeler on your behalf in a Proof of Concept. The consultants will install the product and gather and interpret data on your behalf. This offering, one to four weeks in length, allows you to evaluate the value of Application Workload Modeler in your environment prior to making a purchase decision.
  • A Complex Modeling Offering to assist in complex modeling, providing performance and capacity planning for Enterprise Extender or TN3270. With this offering, you may utilize the skills and machine resources of the IBM team either in conjunction with purchasing Application Workload Modeler (for additional capacity or staffing) or prior to making an Application Workload Modeler purchasing decision.

Further information on services is available here or you can send an e-mail to

Contact IBM
Considering a purchase?
Email IBM

Call IBM:
Priority code:
z Systems