Python Architecture Patterns

14

Profiling

It is quite common that written code doesn't behave perfectly after being tested with real data. Other than bugs, we can find the problem that the performance of the code is not adequate. Perhaps some requests are taking too much time, or perhaps the usage of memory is too high.

In those cases, it's difficult to know exactly what the key elements are, that are taking the most time or memory. While it's possible to try to follow the logic, normally once the code is released, the bottlenecks will be at points that are almost impossible to know beforehand.

To get information on what exactly is going on and follow the code flow, we can use profilers to dynamically analyze the code and better understand how the code is executed, in particular, ...