Building a Cloud-Native SQL Database

on YouTube ↗37m

" summary

Cloud-native SQL database built as a single Go binary that shards data into 64 MB ranges, replicates each range (default three copies) and runs Raft per range for strong consistency; transactions default to serializable isolation using optimistic concurrency with short waits to reduce aborts. System scales linearly (tested to 64 nodes) and hit ~100,000 QPS on YCSB; gossip drives automatic rebalancing, Kubernetes/Helm deploys are provided, and two Jepsen-found bugs were fixed with no remaining known consistency issues.

AI-written summary · about

" chapters (22)

1. 0:00Intro
2. 1:15No sequel databases
3. 3:24Availability and scalability
4. 4:30Choosing a database
5. 4:59Cockroach DB
6. 5:28How the system works
7. 5:44Three main areas
8. 5:59System structure
9. 12:43Range Size
10. 14:49Primary Secondary Replication
11. 16:10Consensus
12. 19:03Raph
13. 20:35Serializable Isolation
14. 21:34Transactions
15. 22:37Simple Example
16. 26:10How to deploy
17. 27:28Questions
18. 27:58Reads
19. 31:30Adding nodes
20. 34:28Migrating
21. 35:23Business Model
22. 36:38Latency

" related talks

• 30mBuilding Cloud Native Applications on Distributed SQL PlatformSHIGERU URUSHIBARA · Distributed SQL Summit Asia 2023
• 42mThe hows and whys of a distributed SQL databaseAlex Robinson · Strange Loop 2017
• 35mA Customer’s Journey: Migrating from NoSQL to a Cloud Native Distributed SQL DatabaseSRIRAM SITARAMAN · Distributed SQL Summit Asia 2022
• 26mHow to Assess Your Cloud Native Workloads for Migration to a Distributed SQL DatabaseMASAKI YAMAKAWA · Distributed SQL Summit Asia 2022
• 26mNewSQL: concepts, realities and when you should embrace itGargi Ahluwalia · XConf Australia 2022
• 52mCockroachDBBen Darnell · Golang UK Conference 2015
• 37mRunning a Cloud-Native SQL Database on MesosBen Darnell · MesosCon North America 2017
• 26mCloud Native Database Meetup Japan: The State of Distributed Database in JapanTAKAHIRO KOBAYASHI · Distributed SQL Summit Asia 2022