Frontend Resilience
summary
BuzzFeed documents concrete fixes not theory: measure real failure rates like ~1% of JS requests timing out and 13 million monthly timeouts, and ~9% users with content blockers. Deploy CSP integrity hashes to freeze third‑party scripts, forward Sentry alerts into Slack for triage, run active-active redundancy across CDNs and image services, keep a static fallback called Plan Z, and use service worker stale serving plus chaos drills and blameless postmortems to discover and fix single points of failure.
AI-written summary · about
chapters (18)
- 0:00Intro
- 1:35RESILIENCE IS FUNCTION IN A HOSTILE ENVIRONMENT
- 2:29UNDERSTAND YOUR TIERS OF USER EXPERIENCE
- 3:06GUARANTEE THE MOST BASIC LEVEL OF UX
- 3:47HOW OUR SYSTEMS FAIL
- 10:41KNOWING IT'S BROKEN BEFORE TWITTER DOES
- 16:20HOPE FOR THE BEST?
- 20:22SOMETHING BROKE!
- 21:47DESIGN FOR FAILURE 1. PRIORITIZE CRITICAL PARTS OF THE PAGE 2. MAKE ERRORS A FIRST CLASS CITIZEN
- 23:25MITIGATE RISK 1. LOCK YOUR RUNTIME DEPENDENCIES
- 28:43MITIGATE RISK LOCK YOUR RUNTIME DEPENDENCIES 2. BUILD IN REDUNDANCY 3. SERVE STALE CONTENT
- 30:03HOW DID WE HANDLE IT AS A TEAM?
- 30:14HOW COULD IT HAVE BEEN PREVENTED?
- 30:57FIRE DRILLS ARE A SAFE SPACE TO PRACTICE
- 31:17DELIBERATELY INTRODUCE FAILURE TO ENSURE YOUR SYSTEMS ARE RESILIENT
- 31:49KNOW WHAT'S IMPORTANT TO YOUR USERS
- 31:59IDENTIFY HOW YOUR SYSTEM WILL DEGRADE
- 32:12LEARN FROM EVERY FAILURE
related talks
32mSurviving Black Friday - A resilience engineering tale
53mResiliency and Availability Design Patterns for the Cloud
28mReliability Lessons from Running High Traffic Services for More Than 10 Years
16mEngineering Failure-Resilient Systems
52mPerformance & Stability Anti patterns
30mAvoiding software fails: few metrics to improve application reliability
42mHow to stay alive even when others go down:...
42mPlanning To Fail