CREATE INDEX … USING VODKA VODKA CONNECTING INDEXES
summary
Prototype of a new PostgreSQL access method called vodka access method implements composable index layers (radix trees plus existing methods) to index jsonb paths and arrays, aiming to replace gin's hashed-path tricks. Tests on NATAS-like dataset show much smaller indexes and faster queries — e.g. document index size 44 MB vs 387 MB and typical lookups become >10x faster compared to naive approaches and competitive with MongoDB. Includes a simple heuristic optimizer for jsonb predicates, supports star-path searches, and is delivered as an extension prototype.
AI-written summary · about
chapters (7)
- 0:00Introduction
- 2:27JSON issues in PostgreSQL
- 9:36JSON search extension
- 13:10JSON indexes in PostgreSQL
- 27:05VODKA access method intro
- 34:14Summary and future plans
- 37:26Conference announcement
related talks
53mAdvanced Indexing Techniques in PostgreSQL
35mLeveraging PostgreSQL 17's JSONB for Enhanced Data Modeling
42mPostgreSQL Tips & Tricks For App Devs
45mA Deep Dive into PostgreSQL Indexing
49mFlexible Indexing with Postgres
45mSQL Indexing Beyond Simple Columns: Expressions, Text Fragments, Attributes J...
46mA Deep Dive Into PostgreSQL Indexing
37mRules of Indexing in PostgreSQL