What I Learned From A Dress, an On-Call Nightmare
summary
BuzzFeed sees 670,000 active visitors hit a viral post and a NoSQL-backed vote collector trips the backend so front-end servers can no longer serve new traffic. Ops is juggling mongo racks and a legacy PERL LAMP app that assumes capacity and doesn’t autoscale, engineers patch a route to accept votes, and the incident exposes technical debt, brittle assumptions, repeated toil and several hours of human cost before a reasonable fix holds.
AI-written summary · about
chapters (8)
- 0:00Intro
- 0:43OnCall Nightmares
- 1:28Llamas
- 1:58Alerts explode
- 2:28The dress
- 2:58The dress phenomena
- 3:28What happened
- 4:58Conclusion
related talks
20mWhat I Learned From A Dress, an On-Call Nightmare
05mIgnite - What I Learned From a Dress an On-Call Nightmare
34mNightmare Before Best Practices
28mReliability Lessons from Running High Traffic Services for More Than 10 Years
41m12 Factors of Pain and Suffering
20mDevOps Without The "Ops" – A Fallacy? A Dream? Or Both?
22mWhat can possibly go wrong? The story of an outrage at Booking.com
30mThe last in the line: a talk about DevOps and Disasters