|
|
General Description
eValid LoadTest enables you to accurately test the scalability and performance
of your Web applications and takes the guessing game out of predicting how your
Web site will handle user load.
eValid achieves server loading by running multiple browsers each with the same, or possibly with different scripts, and measuring times in tester-assigned groupings. To avoid programming you can reuse the scripts created for functional testing without modification and simply apply them in this load performance test. (Read the Detailed Summary.)
Scenario Composition, Reporting
eValid LoadTest scenarios can include any mix of script playbacks,
with data reporting organized according to indvidual user classes.
A LoadTest run consists has these parts:
Master Script. The master script that specifies which script to run and the user class to which it belongs. There can be one script or many scripts, depending on the scenario choices. Repetition Counts, Delay Multiplier. Each LoadTest browser can have its own delay multiplier value, and its own repetition count value. Cache Suppressed. Each script playback runs without cache to assure that pages are actually downloaded from the target server.
Result Reporting, Charting
The LoadTest monitoring screen tracks the progress of each user class
and indicates fastest, slowest and average response times for each defined
user class.
At the end of the test run a chart shows this same information.
In addition, the LoadTest detailed logs are available for further analysis.
Variable Fidelity eValid Playback, eVlite Playback Agent
Normally eValid playbacks are full fidelity and context preserving.
Playback modes in eValid also can serve only text, or only URLs, but without the same fidelity.
The eVlite facility allows ULR only service of standard eValid scripts in up to
1000 simultaneous users per instance,
with up to 10 instances per machine (10,000 simulated users per machine).
Playback Duty Cycle
If a typical user session lasts 10 minutes including "think time",
but takes only one minute of actual playback time
when the playback delay multiplier is set to 0.0,
we say that the
duty cycle for that test is 10%.
Operating a LoadTest scenario that normally has an average 10% duty cycle
at 100% duty cycle will multiply the effective applied load by 10X.
Caution is advised because high duty cycle tests
may impose unwanted saturating loads on local resources.
Machine Capacity
The LoadTest engine is licensed on a per-machine basis.
You can run as many eValid sessions on that machine as it will permit.
On industry standard PCs you can expect 50-150 eValid playbacks
in parallel with normal playback duty cycles.
There are special short-term licenses available to permit running
scenarios on an unlimited number of machines.
