Performance Test Workload Modelling
What is Workload?
Test of Performance The distribution of the load over the defined scenario is referred to as the workload. In order to recreate a real-world scenario in the performance test environment, the performance tester creates a workload. Different workloads are developed during the performance testing cycle to examine the system’s behavior under varied loads and circumstances. The term “workload” can also refer to a circumstance.
What is Workload Modelling?
Workload modeling, sometimes referred to as “performance test scenario creation,” is the process of developing a performance testing scenario using non-functional requirements and performance test scripts.
Purpose of Workload Modelling:
- Workload modeling assists in creating the anticipated production-like conditions in the performance test environment to verify the application’s actual performance.
- Workload modeling also serves the aim of simulating identical conditions to duplicate the issue in the performance test environment when production performance issues are discovered in the live environment.
Accountability:
The preparation of the performance test scenario exactly as specified in the performance test plan is the duty of the performance test analyst or engineer (designation may vary depending on corporate policy).
Approach:
Although there are many calculations involved in workload design, the end goal is to attain the desired goal in a variety of ways. The number of users, transactions per second, hits per second, throughput, etc., might all be included in this list. A predictable (stress test), repeatable (regression load test), scalable (stress/breakpoint test), and long-lasting (endurance test) scenario may be created with the use of the workload model. Gathering all the necessary data is crucial before building a workload model since it enables the development of an efficient workload model. The Performance Test Plan contains this information.
The number of different test types listed in the performance test plan should be examined first. There will be many workload models, one for each kind. As an illustration, the user load in the stress test is more than the user count in the load test. As a result, everyone will have a unique workload model or scenario.
Gather the appropriate metrics now for each of the cited cases. The following are the workload metrics:
- Number of Users
- Iteration per hour
- End to End Response Time (Iteration Response Time)
- Number of Transactions
- Pacing
- Think Time
- Hits (Requests) per second
- Throughput
- Individual page response time
Does a performance tester need all these metrics to design a scenario?
No, to make it easy, understand the category of the workload metrics:
- Number of Users (Basic)
- Iteration per hour (Basic)
- End to End Response Time (Calculative + Targeted)
- Number of Transactions (Calculative)
- Pacing (Calculative)
- Think Time (Calculative)
- Hits (Requests) per second (Targeted)
- Throughput (Targeted)
- Individual page response time (Targeted)
The fact that simple measurements cannot be targeted is not worth mentioning. To give a fundamental knowledge of the workload model, all the metrics in the category were divided. We’ll talk more about the advanced degree of workload later.
Two metrics are crucial for the creation of a fundamental workload model. These two metrics’ values must be included in the performance test plan:
- The anticipated number of users using the server at any given time throughout the test.
- Iterations per second: The number of iterations that must be finished by a user within the test’s length.
Deliverable:
At the conclusion of this stage, a performance tester generates and provides the performance test scenarios. These hypothetical situations may go by titles like Stress Test Scenarios, Load Test Scenarios, and so on.
