Features and Benefits
Highly Scalable
Forecast
TM is capable of simulating high user concurrency without the expense and management overhead of excessive hardware. Forecast
TM can be configured to deploy large numbers of machines during test execution in order to generate extremely high levels of load. The Forecast
TM runtime is very efficient with a low resource footprint.
Comprehensive protocol support
Forecast
TM targets a wide range of technologies including Web/HTTP, Web Services, client-side Java, client-side .Net, Citrix, GUI replay, and network level messaging. Forecast
TM is not a limited "Web only" tool; technologies can be combined to test systems with complex architectures.
Flexible and open
Forecast
TM is modular and extensible allowing applications and IT frameworks to be specifically targeted for even more effective testing.
User-defined Virtual User types include execution runtime extensions, script generation rules and even specialised script generators. The aim is more effective testing whilst reducing the need for manual script editing.
Information not just data
During a load test the thorough checking of server responses and the accurate measurement of response times is vital. Forecast
TM supports nested and asynchronous transactions as well as user-defined metrics. The level and detail of recorded information is configurable to ensure effective problem diagnosis.
ForecastTM integrates data from a range of monitoring sources into both a runtime dashboard and post-run analysis and reporting. This allows the matching of server and network resource utilization with response times and errors.
Intelligent scripting
Creating robust scripts is often the most time consuming and expensive part of testing. Two key problems are data correlation and parameterization. Correlation is where dynamic data returned from the System Under Test (SUT) subsequently forms part of input data sent to the SUT by the Virtual User. A common example of data correlation is the use of a session identifier. Parameterization is where input data is varied to ensure realistic testing. For example, each Virtual User logs on with a unique user identity and password.
ForecastTM is able to record test user journeys and automatically generate scripts. A hierarchy of mechanisms for intelligent script creation is provided to minimise manual editing:
Level 1: Default mechanisms. The script generators and runtime automatically correlate commonly occurring patterns of data and provide easy-to-use methods for parameterization.
Level 2: User-defined generation rules. The script generators allow the user to define additional data correlation and parameterization rules that encapsulate application specific knowledge.
Level 3: User-defined Virtual User extensions and script generator extensions. For unusually complex or non-mainstream applications the standard script generator can be extended to target the application along with extensions to the runtime Virtual User classes.