The NoSQL Way in PostgreSQL
summary
PostgreSQL claims it handles NoSQL workloads by adding hstore, JSON (9.2/9.3) and JSONB with GIN index (9.4) while keeping ACID. A benchmark harness run on Amazon with 50 million records compares PostgreSQL 9.4 to MongoDB 2.6 using COPY for bulk load and single-node, untuned setups. Results show PostgreSQL loads and inserts faster (MongoDB ~270–295% slower in those tests), queries faster (MongoDB ~465% slower), and uses less disk. Node.js examples and a public GitHub tool are available and the benchmark artifacts validate the claim.
AI-written summary · about
chapters (12)
- 0:00NoSQL in PostgreSQL intro
- 1:38Origins and use cases
- 4:04Key-value and document storage
- 7:38JSON data type support
- 9:23JSONB binary format
- 15:27JSON vs JSONB
- 21:13JSON operators in 9.4
- 26:11JSONB vs BSON
- 28:40hstore vs JSON vs JSONB
- 35:49PostgreSQL vs MongoDB benchmark
- 41:51Structured vs unstructured data
- 45:06Resources
