Key Features

Realistic - virtual user based load. Each virtual user is assigned a scenario that he will execute during the test. Between each page request it is possible to use configurable think time. Standard session handling techniques including cookies are supported.

Standard and proprietary protocols – Our technology supports standard protocols such as HTTP and HTTPs but has flexible architecture to support additional protocols.

Flexible user scenarios with dynamic data replacement – Scenario parameterization and data seeding make it possible for each virtual user to supply different dynamic data to the application, such as username, password and other application dependent data.

Distributed load testing controlled by a central system allows for high scalability (1000s of virtual users) and geographically distributed virtual users. This feature enables us to test infrastructure performance from certain points in customers distributed environment.

Response validation – Each response from the application is validated in each step to detect functional and system errors. This is very important when evaluating the error occurrence relative to the load.

Server monitors – Besides providing performance metrics measured on the client side, we constantly monitor critical server resources to detect bottlenecks and use it as input for capacity planning.

Limited bandwidth tests – This feature allows us to understand how the application will perform for users with different types of access – LAN, DSL, Dial-Up, etc.

Unique IP addresses for load balancing testing. Load balance mechanisms should be tested as part of the overall system infrastructure. We use unique IP addresses for virtual users in order to do this.

In-house or in our labs – We can do the testing on your hardware or in our dedicated secure labs. All the testing hardware can be leased to help you reduce the overall investment in hardware.

Comprehensive reports – We focus on providing reports that will highlight most important performance measurements and issues found during the test. However, detailed data is available for in-depth analysis and performance debugging. We will report on page response time, percentage of errors (both system and functional), capacity for active users, workload in transactions per second that application can successfully handle, server resource usage, etc.

Chart Samples – In addition to detailed report figures, we are also providing data charts with clear representation of key response metrics and analysis. Following are some typical charts that represent application behavior seen through the scope of some very important criteria:

Describes the distribution of user commands per response time.
User command is one logical step within a user scenario.

Describes measured number of users that application was able to handle, comparing to projected number
of users defined by a test scenario.

Clearly shows improvement in the application between two test cycles,
since application is capable of handling greater number of requests
per second during the second test cycle.

Shows application errors during the duration of a 12 hour load test.
Most of the errors happen within first 3 hours, since during that
time most of the users are logged in and at that time they begin
to die out gradually.