6.

When a client participates in a benchmark run, it simulates a user by following a script. An example of such a script is shown in Figure 5: Benchmark scripts. One script contains a number of steps. Some of the steps relate to real user actions, such as “create sales order”, and other steps contain script basics- such as loop, end loop, sales order, etc. Steps usually performed by real users are marked as a dialog step. For each dialog step, an expected response time in seconds is given. Note that steps performed frequently have a smaller expected response time.

If the script runs at full throttle, it can run many times faster than a real client would. To compensate for this delay, each dialog step can be set-up so that a “think time” is added.  By adding think time, an upper limit is added to the possible throughput.

The benchmark tool allows throughput to be measured in a certain amounts of units. This is specified on the script by flagging one of the steps with the “line indicator”. The throughput is then simply calculated as the number of times the line indicated step is executed over a certain amount of time.

The script usually has two implementations; one that is display dependent, using forms, and one that is non-display dependent. These can be seen on the Benchmark Run form.

@%20=58725AB5=.files/image005.jpg">

Figure 5: Benchmark scripts.

Configurations

When a client participates in a benchmark run, it simulates a user by following a script. An example of such a script is shown in Figure 5: Benchmark scripts. One script contains a number of steps. Some of the steps relate to real user actions, such as “create sales order”, and other steps contain script basics- such as loop, end loop, sales order, etc. Steps usually performed by real users are marked as a dialog step. For each dialog step, an expected response time in seconds is given. Note that steps performed frequently have a smaller expected response time.

If the script runs at full throttle, it can run many times faster than a real client would. To compensate for this delay, each dialog step can be set-up so that a “think time” is added.  By adding think time, an upper limit is added to the possible throughput.

The benchmark tool allows throughput to be measured in a certain amounts of units. This is specified on the script by flagging one of the steps with the “line indicator”. The throughput is then simply calculated as the number of times the line indicated step is executed over a certain amount of time.

The script usually has two implementations; one that is display dependent, using forms, and one that is non-display dependent. These can be seen on the Benchmark Run form.

@%20=58725AB5=.files/image005.jpg">

Figure 5: Benchmark scripts.

Configurations