~/talk

Finding a Needle in the Haystack - Debugging Performance Issues

Talkat PyCon India 2025 · 2025-09-12..09-15· [testing]
slides:none on file

Practical debugging steps for Sev1 slowness using Python profilers with concrete results: deterministic profilers (cProfile) and sampling tools (py-spy, pyinstrument) plus memory tools (memory_profiler, tracemalloc, mprof, Memray) and flame graphs. Shows how to attach to PID and get a hot line without restarts, how replacing string concatenation with join cuts time and memory, a real incident where a regex running a million times was the culprit, a boot-up delay from an accidental class variable, and a costly double marshalling that spikes GC.

AI-written summary · about

  1. 0:00Introduction and speaker intro
  2. 1:47Speaker background and context
  3. 3:12Performance issues analogy
  4. 5:21Importance of performance testing
  5. 6:27What is profiling
  6. 8:37Deterministic profilers
  7. 11:31Sampling profilers
  8. 14:06Memory profiling importance
  9. 17:00Memory profiling tools
  10. 20:41Garbage collection impact
  11. 23:39Real-world debugging examples
  12. 27:59LLMs and debugging context
  13. 28:43Key takeaways summary
  14. 29:26Final advice and credits
  15. 29:50Q&A session
~/search
type to search 6,518 conferences, 143,918 speakers and 192,133 talks
~/help
cmdctrl k
search every conference, speaker and series
/
focus the filter search
h
home
c
conferences
s
speakers
e
series
t
light / dark — the status bar button does the same
← →
on a talk: 5 seconds back / forward
0–9
on a talk: jump to 0%–90% of the recording
?
this help
esc
close this, then clear every filter — or just the search, while you are typing in it

Ignored while you are typing. Pages without a filter rail have no search to focus.