Load Test

Load testing measures the website's (through the http(s) service) capabilities. We can launch load test agianst other services as well. Measuring the performance of code under duress. The Performance measures include several parameters such as: Elapsed time, Availability, Total data transferred, Server response time, its Transaction rate, its Throughput, its Concurrency and the number of Times it returned normal.

Tasks to perform a Load Test

  • Create scenarios consisting to simulate the action of real users through the site.
  • Validate the scenarios.
  • Executing tests and collecting results.
  • Analyzing the results.
  • Issuing the report.

  • Types of Load Test

    There are multiple types of load test in which every type provides different results. We mention:

    1. Degradation test (Breakpoint test)

    It is a technical test during which the analyst will simulate activity on the website, in order to determine how much simultaneous load the system is capable of supporting for each function of the application. This helps determine the maximum number of concurrent users that the site is capable of supporting. Therefore, the goal of this type of test is to test the “Break Point”, the one which will give the maximum load supported by the application.

    2. Performance test

    This is a test during which the analyst will measure the performance of the application subjected to a user load. The nuance with the previous type lies in the fact that we seek to verify the intrinsic performances at different levels of user load. This helps determine the optimal level of activity supported by the website. Therefore, We seek to verify the hypotheses of the degradation test analysis and validate the intrinsic performances at different load levels. The ideal is to have access to the server during this test to measure the memory consumed, the use of the CPU, the SWAP if there is one, the use of the bandwidth

    3. Robustness test

    This is a test in which the analyst simulates a large user load (that the website is capable of supporting) over a long period of time, to see if the system under test is capable of support intense activity over a long period without degradation of application or system resources. Therefore, this test simulates an optimal number of users (the number above which the user experience becomes non-optimal) over a relatively long period (simulate activity for days/several days/months/years, etc.).