performance-engineering

Profile Python web services by attaching py-spy to a running process to generate CPU flame graphs, and use memory_profiler to identify memory bottlenecks. Run `py-spy record --pid <PID>` to output visual profiling data.

Yes, you can use Locust and k6 for load testing Python web applications. Both tools simulate real-world traffic to measure performance under load, helping identify I/O and database bottlenecks before they impact production.

Optimize database queries in Python web services by profiling slow queries, applying connection pooling best practices, and identifying I/O bottlenecks through load testing. This reduces latency and improves overall application throughput.

Yes, you need Python and py-spy installed to execute performance profiling scripts. The Skill requires Python, py-spy, Locust, and other related tools to be present in the environment to run tests and optimize web services.

Load testing helps scale WebSocket connections by simulating concurrent traffic with tools like Locust or k6. This measures performance limits under load and highlights I/O bottlenecks, enabling effective WebSocket scaling strategies.

Use py-spy for CPU performance optimization to visualize thread execution and find CPU bottlenecks, and use memory_profiler to monitor memory usage line-by-line. Both are needed to address different resource bottlenecks in Python web services.