Why do you perform performance tests before the software goes online? How to find a software performance testing organization

We can expect , They always react , Let's look at many examples from the past , Explain how they can proactively identify performance issues before going live , To make better use of software components and improve the customer experience .
Of course , And “ Invest more hardware to solve application performance problems ” Compared with , The benefits of proactive performance planning far outweigh this approach , But how relevant is this practice in today's era ？

 Software delivered DevOps Method , Expect to deliver new functions at supersonic speed ？
 Cloud adoption can be achieved through simple strategies / Rules to configure additional hardware
 Can control the service 、 New deployment strategy for server and user deployment
 We cannot build a test environment at the right scale or scale to reflect the complexity of the application ecosystem
 New observability tools can help speed up diagnosis , So that the root cause can be quickly identified , And DevOps When combined , You can quickly push and test ways to deliver changes in a production environment rather than in a test environment
Implemented in their process CI/CD After pipeline , Some customers also question the need for performance testing , Because they can release any fixes to the production environment when they occur . When they can solve problems in production faster than before , Why invest in active performance testing ？ As applications are also migrated to the cloud , When we can take op-ex Routes effectively maintain our applications in the cloud , Whether it will bring any value by pitching in early performance tests ？
When is the best time to do software performance testing ？
1. Start early
2. Before going online
3. Move right ： After the launch
4. Move right ： Use the observability platform to RCA And fix the problem
We don't think there is a magic bullet to reduce performance risk .
The answer needs to be through consideration of organizational background and culture 、 Budget expenditure and risk appetite 、 Business criticality and brand identity 、 Delivery practice maturity 、 Technology uses curves and market introduction schedules to determine .
All these factors will determine how the organization responds to “ In the early ” Test and “ Delay ” To test or “ No test ” And “ Quick diagnosis and repair ” Test this new challenge , Instead of investing in long-term performance testing before going online .
There is absolutely no better way than another . You may have to choose a context specific method , This approach requires performance engineering practices to balance risks 、 Time and cost .