Two Years with Vitess: How JD.com Runs the World's Largest Vitess
summary
JD.com runs the world’s largest Vitess on Kubernetes deployment, adopted in 2017 to scale MySQL and cut machine and dev cost. Production footprint hits 20,000 DB tablets across hundreds of shards, automated resharding routes traffic in seconds and uses two-phase commit to avoid partial commits. Built components include a custom orchestrator (OFC), etcd-backed watch routing, a JTransfer real-time extractor into big data, and ML-based scheduling; ongoing work is migrating clusters to Vitess 3.0 and tightening monitoring and failover handling.
AI-written summary · about
chapters (8)
- 0:00Database sharding concepts
- 5:49Sharding and commit mechanics
- 9:41Managing multi-shard transactions
- 13:25Secondary indexing strategies
- 16:03Internal architecture design
- 21:08Business workflow and automation
- 25:55Best practices for users
- 28:57Future plans and challenges
related talks
35mTwo Years with Vitess: How JD.com Runs the World's Largest Vitess
36mRunning Vitess on Kubernetes at Massive Scale: JD.com Case Study
43mVitess and a Kubernetes Operator
40mVitess: Intro and New Features
45mDeep Dive: Vitess
05mRun Your Data Layer In K8s With Vitess
31mVitess: Running MySQL with confidence on Kubernetes
05mVitess: Unlimited Database Scalability