Using FoundationDB and the FDB Record Layer to Build CloudKit
Talkat FoundationDB Summit 2019 · on Monday, 2019-11-18· at 11:40· Grand Ballroom 5, Marriott Marquis San Diego Marina·
summary
CloudKit runs on FoundationDB and the FDB Record Layer and uses per-app, per-user contiguous record-store key ranges so millions of databases move and rebalance by range splits; multi-key ACID transactions enable advanced secondary indexes including a version-stamp sync index that gives globally ordered, conflict-free modifications; the Record Layer implements stateless, streaming queries with opaque continuations and resource limits; transactional, low-overhead per-record-store text indexes replace external search services; result: billions of databases at high availability and faster, safer upgrades.
AI-written summary · about
chapters (16)
- 0:00Intro
- 1:14Thousands of Apps, Billions of Databases
- 1:51FDB, Record Layer, and Record Stores
- 4:13Record Stores and Sharding
- 4:32Balancing the Load
- 5:33Record Layer for Stateless Compute
- 7:01Streaming Queries
- 8:11Limits, Throttling, and Resource Management
- 9:47CloudKit and Indexing
- 10:31High Concurrency Sync
- 12:15Record Sync - The Old Way
- 14:19Record Sync - Everything's Better with Versionstamp
- 15:03Record Sync With Versionstamp
- 16:20Text Indexes
- 17:44FoundationDB Robustness in Practice
- 19:04Summary
related talks
46mFoundationDB: The Best-Kept Secret of New Distributed Architectures!
37mTechnical Overview of FoundationDB
25mFoundationDB Document Layer
50mBuilding the Next-Generation Backend on FoundationDB
22mSolving Everyday Data Problems with FoundationDB
34mFoundationDB, the Black Knight
22mHow FoundationDB powers SnowflakeDB's metadata!
53mBuilding Distributed, Multi-Tenant Databases with FoundationDB and Rust