Because performance analysis is not always a part of software engineering or computer science courses, many software engineers cannot deal with performance optimization well. In a large company like IBM, they will hire professionals who have received special training in dealing with software performance. If the company develops software applications with more than 10000 function points, such experts are certainly the best candidates. At present, there are many performance tools and measurement devices, such as analyzers that collect dynamic data. Of course, it is also feasible to embed the measurement capability of software performance into the software application itself, which is known as "piling technology"
Since pile insertion technology and other forms of performance analysis tools may reduce the running speed of the program, necessary maintenance is required to ensure the correctness of the data. Several physical terms are used in the field of performance optimization. For example, "Heisenbug" is named according to the Heisenberg uncertainty principle. It refers to the vulnerability that disappears every time when it is read and analyzed. Another physics term is "Bohr bug", which is named after Niels Bohr. This kind of bug can only appear under special conditions. The third physics term is "Mandel bug". It is named after Bernet Zedelbrodt, who created the chaos theory. This form of vulnerability is caused by random and chaotic factors, so it is difficult to isolate it. The fourth vulnerability is "Schroden bug", which is very rare and named after Ernst Schroder. Normally, this vulnerability will not occur unless someone realizes that those codes should not be executed at all; And it is said that when these vulnerabilities are found, the software will stop working.
The emergence of performance problems is also based on the business cycle. For example, many financial and accounting software packages will run at a significantly slower speed after being used in a quarter or a fiscal year, which is due to the sudden increase in usage. When the software is running, if serious vulnerabilities cause the software to fail to run, the performance of the software will be zero. However, there is no good discussion on it in the performance chapter. This problem can be measured by mean time to failure. Such problems are common within one to two months after software delivery, but will decrease with the gradual stability of the software. It is increasingly common for denial of service attacks to cause software to stop working.
Finally, the best practice of performance is similar to that of quality control and safety control. A common best practice is to hire performance professionals and excellent people in quality control and safety control. In terms of security, the effect of AI or neural network performance optimization is much better than that of testing or artificial performance testing experts in finding performance. A similar software can be used in architecture and design. This software can provide the best performance optimization scheme according to performance optimization rules and algorithms before coding.
Generally speaking, artificial intelligence and neural networks are recommended for dealing with complex problems, such as security vulnerabilities and performance problems. These topics cross autonomic computing or applications, and these applications tend to monitor and improve the performance and quality of the local computer.
