Y2K and Other Disappointing Disasters: Risk Reduction and Harm Mitigation
summary
Y2K required thousands of engineers to edit mainframe date logic and reboot servers that sometimes never came back, and minor July outages still occurred; examples include BitLocker adoption, Halon data centers, the Delta data center fire and the 35W bridge collapse where multiple small failures combine. Practical mitigations cited are canary launches and feature flags, kill switches and circuit breakers, Gremlin/Simian-style chaos testing, data duplication, and status pages off your main storage (S3 outage showed why).
AI-written summary · about
chapters (16)
- 0:00Y2K & Other Disappointing Disasters
- 7:56Risk Reduction
- 8:46How to reduce risk
- 12:15Accept Risk
- 14:26Harm Mitigation
- 21:08Fail safe or fail secure?
- 23:40Assume failure
- 25:39It's complicated and hard
- 28:44What's a disaster?
- 33:23Microservices and loose coupling
- 34:09Fault-tolerant messaging
- 34:56Data duplication
- 36:32Canary launches
- 37:40Kill switches and circuit breakers
- 38:30Automatic recovery
- 38:56Testing for load, stress and outage
related talks
51mProgramming's Greatest Mistakes
01h 04mLocknote: Programming’s Greatest Mistakes
34mProgramming’s Greatest Mistakes
22mSemi-Advanced Threats vs the Semi-Critical Infrastructure
49mEverything is broken and that's Okay!
54mThe Error of our Ways
30mFueling The Future: Building Robust Engines
44mFailure Is Always An Option