Performance testing methods can be divided into two types: static and dynamic.
Static performance testing
Static performance testing (hereinafter referred to as static testing) is often more important than functional testing in performance testing, because many serious performance and efficiency defects are introduced during the system architecture design phase, such as unreasonable or unbalanced system architecture. Problematic algorithm models, etc. These defects may be introduced due to misunderstandings or lack of relevant knowledge by designers and architects, or because design requirements do not adequately capture factors such as response time, throughput rate, resource utilization goals, expected load and usage, constraints, etc. Therefore, static testing is particularly suitable for carrying out in the early stages of system construction.
Static testing includes:
❑Requirements review on performance and performance risks;
❑Review of database architecture, entity relationship diagram, metadata, stored procedures, queries, etc.;
❑Review of system and network architecture;
❑Review of code for key parts of the system (such as complex algorithms).
Dynamic performance testing
When the system has been built, dynamic performance testing (hereinafter referred to as dynamic testing) should start as early as possible. Several key testing opportunities are as follows.
❑During unit testing, use information analysis to identify potential bottlenecks and dynamic analysis to assess resource utilization.
❑During component integration testing, test key use cases and workflows across components, especially when integrating different use case functionality or integrating with the backbone structure of a workflow.
❑During system testing, check the overall end-to-end performance under different load conditions. ❑During system integration testing, especially when testing data flows and workflows at interfaces between critical systems, the “user” may be another system or machine (such as a sensor or other input system).
❑In the user acceptance testing phase, build users, customers, and operators' confidence in the system's performance, and tune the system under real conditions (but this is usually not to discover performance defects in the system). If custom or new hardware is part of the system, you can use an emulator to perform early dynamic performance testing. However, it's best to start testing on real hardware as soon as possible, since simulators often do not adequately capture resource constraints and performance-related behavior.
In higher-level testing such as system testing, system integration testing or user acceptance testing, the use of real test environments, data and loads is critical to the accuracy of performance test results.
In agile and other iterative and incremental development models, teams should incorporate static and dynamic testing into early iteration plans for the software rather than waiting until the final iteration to start performance testing.
This article "Enterprise Performance Testing"