Chef @ Yahoo! JAPAN: How Chef Powers the...
summary
Claims that Chef runs Yahoo! JAPAN’s infrastructure at scale: a single cluster managing 53,000 servers across 100+ services, with ~23,000 chef-client runs/day. Pipeline consolidation moves from a 50+ Jenkins sprawl to screwdriver plus Chef, cookbooks replace fabric playbooks, and ad systems run ~2,000 nodes now deploying 100 servers concurrently (planning 4,000). Pain points are cookbook quality, yearly upgrades across tens of thousands of nodes, and backend HA/BCP needs.
AI-written summary · about
chapters (14)
- 0:00Start
- 1:04Introduction
- 5:43Financial Services
- 8:59Mapping Services
- 11:15Yahoo Installer
- 22:13User Interface
- 22:23Steps as a Pipeline
- 23:32Integrated Configuration Management Methodology
- 24:44Lack of Chef Skills
- 28:40User Support
- 30:46History of Server Nodes Growth at Yahoo Japan
- 33:09Benefits of Bringing In Chef into the Chef Infrastructure
- 36:35Ui Improvement
- 36:54Bcp Business Continued Continuity Planning
related talks
25mManaging Changes Seamlessly on Yahoo's Hadoop Infrastructure Servers
16mCase Study: How Yahoo! JAPAN Uses and Contributes to Screwdriver at Scale
25mCooking with Chef - Your servers will thank you
33mScaling Chef Emotionally
40mJenkins and Chef: Infrastructure CI and Application Deployment
33mDeveloping for Multi-Component Environments While Keeping Your Sanity
05m3+ Useful Ops Tools in Japan
01h 01mOpening Keynote