TechEmpower.FrameworkBenchmarks/frameworks/Java/spring-webflux/README.md

Spring Webflux Benchmarking Test

This is the Spring Webflux portion of a benchmarking test suite comparing a variety of web development platforms.

Netty is used for the async web server, with nearly everything configured with default settings. The only thing changed is Hikari can use up to (2 * cores count) connections (the default is 10). See About-Pool-Sizing

A fixed thread pool of size equals to the number of database connections is used to run all the blocking code (jdbc database accesses) to not block netty's event loop.

For postgresql access, there are four implementations.

• JdbcDbRepository is using JdbcTemplate.
• PgClientDbRepository is using reactive-pg-client
• RxJava2DbRepository is using rxjava2-jdbc
• R2dbcDbRepository is using r2dbc-postgresql For mongoDB access, spring-data-mongodb with reactive support is used. See MongoDbRepository

Plaintext Test

• Plaintext test source

JSON Serialization Test

• JSON test source

Database Query Test

• Query test source

Database Queries Test

• Queries test source

Database Update Test

• Update test source

Template rendering Test

• Template rendering test source

Versions

• Java OpenJDK 10
• Spring boot 2.1.0.RELEASE
• Spring data mongodb 2.1.0.RELEASE
• reactive-pg-client 0.10.6
• rxjava2-jdbc 0.2.0
• r2dbc-postgresql 1.0.0.BUILD-SNAPSHOT

Test URLs

Plaintext Test

http://localhost:8080/plaintext

JSON Encoding Test

http://localhost:8080/json

Database Query Test

http://localhost:8080/db

Database Queries Test

http://localhost:8080/queries?queries=5

Database Update Test

http://localhost:8080/updates?queries=5

Template rendering Test

http://localhost:8080/fortunes