Debugging Python in Production: Practical Techniques Beyond Print Statements
summary
Concrete debugging recipes for production Python: use structured logging with request IDs, timestamps, and severity so errors locate the failing request; correlate stack traces and runtime inspection to find the exact failing line; reproduce issues with small, runnable Python programs the uploader promises and walk through fixes; instrument services with Prometheus and Grafana plus an error aggregator like Sentry for alerts and traces; use tracing to follow a request across services, inspect memory/CPU spikes, and verify whether rolling back or clearing leaked resources actually fixes the problem.
AI-written summary · about
chapters (8)
- 0:00Introduction
- 1:29Debugging production challenges
- 2:11Logging for debugging
- 3:17Tracebacks and exceptions
- 4:41Monitoring system health
- 6:50Tracing requests
- 8:13Observability in distributed systems
- 9:36Q&A
related talks
26mow to Enjoy Debugging in Production with Karen Tracey
31mTracing, Profiling & Debugging in Production (eBPF)
34mDebugging python applications inside k8s environment
41mDetective Hat: Investigating Production Issues
11mUsing and abusing Python's logging module for more readable logs
41mAdvanced Python Debugging Techniques Using GDB
40mProduction Detectives: Solve the Case Before the Crash
18mDebugging Stories from an Advanced Python Class