1.1 system performance is the study of the entire system, including all hardware components and the performance of the entire software stack system is a transaction that requires the participation of many personnel, includes system administrators, technical support staff, application developers, database administrators, and network administrators. 1.2. The performance field includes the following: 1. Set performance objectives and establish performance models; 2. summarize performance characteristics based on software or hardware prototypes; 3. analyze the performance of development code (software integration ). before) 4. perform non-return software tests (before or after software release) 5. Benchmark Testing for software release 6. Conceptual verification in the target environment 7. Configuration Optimization for production environment deployment 8. Monitoring of software running in the production environment 9. Performance analysis of specific problems 1.3. server load analysis and resource analysis system administrator> resource analysis application developers> server load analysis 1.4. Complexity bottlenecks are often complex, they will also contact each other in unexpected ways. Fixed a problem that may only push the bottleneck to other places in the system, resulting in the overall performance of the system not being improved as expected. Therefore, a global approach is often required to solve the problem of complexity. The entire system, including internal and external interactions, may be investigated and studied one by one. 1.5. The most suitable metric for latency and dynamic tracking-latency (used for waiting time) is very useful but not always available. Some systems have average latency, some systems do not have any latency. Dynamic Tracing can measure the latency from any point of interest. It can also provide data to show the complete distribution of latency. This is commonly used in dynamic tracing tools: dtrace.
Performance 1-system performance