Next Generation Data Storage with CouchDB
summary
CouchDB stores native JSON documents with built‑in revisioning and MVCC so reads never see partial writes, serves everything over an HTTP REST API, and indexes queries with map/reduce views (example: a two‑line map plus reduce that produces tag counts). Replication is bidirectional and can sync a full DB to a laptop or eventually sync view subsets; conflicts become separate revisions to resolve. Implementation uses Erlang for massive concurrency, supports attachments, Lucene for full‑text, has compaction, and is labeled alpha but running in production.
AI-written summary · about
chapters (46)
- 0:00Intro
- 0:35I can do that
- 0:54Who am I
- 1:28The basic premise
- 2:35What is CouchDB
- 3:36What is a document
- 5:19Object Notation
- 6:01Deserialize
- 7:02JSON
- 8:57Serialization
- 9:44Documents
- 10:04HTTP
- 11:20HTTP Tools
- 11:39Dump Object Store
- 12:10Views
- 13:26MapReduce Example
- 15:07Map Function
- 16:44View Index
- 18:30View URI
- 19:02Replication
- 22:47Distributed Replication
- 25:58Consistency
- 26:32Build for the future
- 27:11concurrency
- 28:05tradeoffs
- 29:35CouchDB Erlang
- 38:02CouchDB Gracefully
- 38:57CouchDB Consistency
- 40:24Bonus Features
- 41:18Full Text Search
- 42:35History of CouchDB
- 43:56Apache CouchDB
- 44:54Whats next
- 46:01Lightning Talks
- 47:16CouchDB Status
- 47:59Change to two documents
- 49:08JSON structure
- 50:04View
- 50:41Create
- 51:31Object Database
- 52:22Relational Database
- 54:36Question
- 55:19Security
- 56:06Object View
- 58:01Browser Integration
- 59:15Data Types
