How do we know it's working? Proving adblockers production health
summary
eyeo monitors extension health with a four‑layer stack: filter crawler and validator, Sentry and surgical logging, feature flags and user reports. A DNR investigation finds a data mismatch within 24 hours: 60% mismatch between restored and expected rules and 1% with 30k extra rules that saturate browser quotas. A wipe‑and-recreate fix rolls out behind a flag, succeeds for 80% then after race‑condition fixes is validated on 62,000 users and deployed to 100%.
AI-written summary · about
chapters (11)
- 0:00Introduction of speaker
- 1:06Defining working software
- 3:38Key metrics for adblockers
- 5:07Challenges in browser extensions
- 6:30Four monitoring layers
- 8:14Filter monitoring tools
- 10:47Sentry for extension monitoring
- 11:50Custom logging system
- 12:32Case study deep dive
- 17:59Conclusion and mission
- 19:07Q&A session
related talks
52mAB-tests, lies, damned lies, and statistics.
30mBeyond the Dashboard: Build End-to-End O11y
06mWho Am I? A Design Probe Exploring Real-Time Transparency About Online and Offline User ...
33mFrontend Resilience
28mFind friction in your developer experience by user testing with developers
53mHow our critical systems fail, and what we do about it
18mProduct Reliability in
32mCustomer Centric Monitoring