Continuous Performance Improvement of HTTP API

In <a href="https://towardsdatascience.com/how-we-optimized-python-api-server-code-100x-9da94aa883c5" rel="noopener" target="_blank">my previous post</a>, I detailed a few code tricks to improve backend performance. How did I know where to focus and what to optimize, though? Indeed, joining Cython and other low-level gizmos to the party should have solid reasoning. I work at <a href="https://athenian.com/" rel="noopener ugc nofollow" target="_blank">Athenian</a>. Athenian offers a SaaS that helps engineering leaders build a continuous improvement software development culture. We have pretty strict performance targets dictated by the UX. It’s hard to achieve great P95 response times without proper tooling. So we’ve wrapped ourselves with high-quality apps and services: <ul> <li><a href="https://docs.sentry.io/product/sentry-basics/tracing/distributed-tracing/" rel="noopener ugc nofollow" target="_blank">Sentry Distributed Tracing</a> allows us to investigate why a particular API request executed slow in production. This tool works in the Python domain.</li> <li><a href="https://prodfiler.com/" rel="noopener ugc nofollow" target="_blank">Prodfiler</a> gives an independent zoom into the native CPU performance, including all the shared libraries.</li> <li><a href="https://github.com/benfred/py-spy" rel="noopener ugc nofollow" target="_blank">py-spy</a> is an excellent low-overhead Python profiler by  <a href="https://medium.com/u/de31421e853a?source=post_page-----86290433aa54--------------------------------" rel="noopener" target="_blank">Ben Frederickson</a> .</li> <li>Prometheus + Grafana help to monitor the immediate situation and trigger performance disaster recovery.</li> <li><a href="https://cloud.google.com/logging/docs/logs-based-metrics" rel="noopener ugc nofollow" target="_blank">Google log-based metrics</a> augment the previous toolchain by indicating an elevated frequency of important operational events.</li> <li><a href="https://cloud.google.com/blog/products/databases/get-ahead-of-database-performance-issues-with-cloud-sql-insights" rel="noopener ugc nofollow" target="_blank">Google Cloud SQL Insights</a> is a must-have managed PostgreSQL performance monitor on the individual query level.</li> <li><a href="https://explain.tensor.ru/" rel="noopener ugc nofollow" target="_blank">explain.tensor.ru</a> is my favorite PostgreSQL execution plan visualizer. It offers many automated hints that are always relevant.</li> </ul> Today’s post illustrates the weekly routine to identify slow spots and speed up things. In particular, I will demonstrate the usefulness of Sentry traces, py-spy, and Prodfiler. <a href="https://betterprogramming.pub/continuous-performance-improvement-of-http-api-86290433aa54">Read More</a>