maxime-rainville/silverstripe-graphql-benchmark

Create fake Graphql schema to benchmark sizes

dev-master 2022-03-31 09:23 UTC

Requires

• fzaninotto/faker: ^1.7
• silverstripe/cms: ^4
• silverstripe/framework: ^4
• silverstripe/vendor-plugin: ^1
• symbiote/silverstripe-queuedjobs: ^4.9
• symfony/filesystem: ^5

BSD-3-Clause

README

The purpose of this module is create big ass GraphQL schemas to benchmark GraphQL v4 performance.

Installation

composer require maxime-rainville/silverstripe-graphql-benchmark

Usage

The schemas and related DataObject are created by running this command.

# The fake schemas are created against this repo. Make sure you remove any pre-existing schema first
rm vendor/maxime-rainville/silverstripe-graphql-benchmark/src/Models/*.php
rm vendor/maxime-rainville/silverstripe-graphql-benchmark/_config/graphql.yml
rm vendor/maxime-rainville/silverstripe-graphql-benchmark/_graphql/* -R

sake dev/tasks/MaximeRainville-SilverStripeGraphQLBenchmark-Tasks-SchemaCreatorTask "" flush=all

Look at the config option on SchemaCreatorTask for the various options.

If you want to be able to deploy this somewhere, you'll have to run the task and commit your changes to a fork.

What about data

The dummy data object will try to generate some random data and random relationships on dev/build.

By default, we'll generate 5 instances of each DataObject class and randomly link them up in relationships.

What schema will be generated out of the box?

By default, the module doesn't have any schemas or DataObjects. You have to run the task to get something.

By default, the task will generate the following schemas:

• tiny: 5 DataObjects
• small: 10 DataObjects
• medium: 40 DataObjects
• big: 100 DataObjects
• verybig: 250 DataObjects
• gigantic: 500 DataObjects

The generated DataObject will have a mixed of unversioned DataObjects, versioned DataObject and versioned-unstaged DataObjects. The ratio is 45-45-10.

What other branches can I install?

• generic provides a generic set up that should be suitable for most cases.
• small-only only ships a tiny and small schema. Use this branch if you don't want to be stuck building the full 500 DataObject gigantic schema
• versioned-all will add the Versioned extension to all the DataObjects
• versioned-none will not apply the Versioned extension to any DataObjects