Engineering for Engineering's Sake
summary
Company builds a social feed using Redis and hits operational and knowledge gaps that take about a year and a half to resolve, ultimately migrating to Redis Cluster; a similar team at Etsy uses memcached, ships faster and never pays that cost. Core claim: adding new tech for novelty introduces distributed knowledge, ops work and long cleanup; prefer memcached was "good enough" over shiny choices and focus on shipping not just playing with Redis.
AI-written summary · about
chapters (17)
- 0:00Start
- 6:15HOW DO I ENGINEER?
- 7:38A MAGPIE
- 8:03I LOVE TO DO ENGINEERING FOR ENGINEERING'S SAKE
- 12:09HOW TO THINK ABOUT ENGINEERING?
- 14:22REDIS IS A PERFECT FIT!
- 17:12MIGRATION TO REDIS CLUSTER SEPTEMBER, 2015
- 23:04TOOLS
- 25:11LET GOOGLE/FACEBOOK/AMAZON FIGURE IT OUT
- 26:08KNOWLEDGE
- 27:13WAIT, WHAT SHOULD I USE AGAIN?
- 28:01OPERATIONS
- 29:22DON'T ADD MORE UNLESS YOU REALLY REALLY REALLY REALLY NEED IT
- 32:52THE GOAL IS EXCITING PRODUCT
- 33:29THE "RIGHT" TOOL FOR THE JOB
- 34:13USE BORING CODE, TECH, ENGINEERING (AND EXCITING, SOMETIMES)
- 34:27OPTIMIZE COMPANY-WIDE SOLVE PROBLEMS WITH WHAT YOU ALREADY HAVE
related talks
44mEngineering for Engineering's Sake
46mScaling Pinterest
41mHopelessness and Confidence in Distributed Systems Design
26mBe Awesome By Being Boring
25mEgoless Engineering
27mEngineering Productivity Starts with Predictable Behaviour
31mBuilding new on top of old: the argument for simplicity
55mA Decade of Progress: The Evolution of Production Engineering and Reliability