What I Learned From A Dress, an On-Call Nightmare
summary
A viral image drives 670,000 active visitors to BuzzFeed and the monolithic LAMP front end with MongoDB pixel tracking and memcache exhausts connections so editorial voting breaks. The emergency response manually adds servers and tunes the NoSQL-backed vote collector rather than autoscaling; capacity assumptions fail, team toil spikes, and remediation restores service enough for roughly ~10 million votes. Postmortem stresses tech debt and human cost.
AI-written summary · about
chapters (4)
- 0:00Start
- 4:45These issues happen to us all...
- 11:00What was "the dress" phenomena?
- 19:02What's your nightmare?
related talks
06mWhat I Learned From A Dress, an On-Call Nightmare
05mIgnite - What I Learned From a Dress an On-Call Nightmare
28mReliability Lessons from Running High Traffic Services for More Than 10 Years
42mAvoiding Microservice Megadisaster
41m12 Factors of Pain and Suffering
34mNightmare Before Best Practices
10mSurviving the 2008 Elections
30mAbove the Clouds: Investigating Systems from 32,000 ft.
