GraphQL. Accelerated
summary
GraphQL on Node.js becomes fast by concrete changes: use mercurius with an LRU cache to avoid reparsing, benchmark with node clinic and autocannon, and JIT-compile hot queries with graphql-jit to generate JavaScript that V8 optimizes. Batching loaders remove N+1 DB requests so a nested query hits a handful of DB calls instead of one-per-item, pushing latency and throughput near REST levels in benchmarks. The implementation exists and the optimizations work.
AI-written summary · about
chapters (6)
- 0:00Start
- 3:28Nonstandard
- 9:00Makes prototyping easy!
- 12:34On each and every request the server first has to parse the query to produce an Abstract Syntax Tree and then walk that to execute the query
- 14:40mercurius uses an LRU to cache stable queries
- 16:36On the fly code generation
