Fauna Driver Performance Test Utilities

This repo contains common files that Fauna employs to run performance tests across our drivers so we can monitor for performance regressions. The repo is primarily intended for use by Fauna employees.

init.sh and teardown.sh

These two shell scripts handle setup and teardown of the performance test database. Before running these files, the runtime environment must meet these prerequisites:

• Required environment variables:
  • FAUNA_ENDPOINT (the full URI to Fauna, e.g. https://db.fauna.com)
  • FAUNA_SECRET (a database key with write privileges)
• Fauna CLI must be installed (see documentation)
  • npm install -g fauna-shell@^3.0.0

fauna/main.fsl

This file contains Fauna Schema Language (FSL) that defines the two Collections used in the performance tests, Product and Manufacturer. This database schema is "pushed" to the test database using the Fauna CLI fauna schema push command by the init.sh script. For more information on managing schema with FSL, see our documentation.

fauna/seeddata.fql

This file defines a Fauna Query Language (FQL) query that is run using the Fauna CLI to prepopulate the test database's Collections with documents. This FQL file is called during setup in the init.sh script after the database schema is ready. For more information on querying Fauna with FQL, see our documentation.

queries.json

This file contains a JSON object that defines the various performance test queries and metadata; this object gets consumed by the language-specific performance test cases and drives the inputs for the test scenarios. The top-level field, queries, is an array of objects with the following fields:

• name - the name of the test (e.g. basic_add)
• parts - an array of strings that compose the query
• response.typed - whether the query result is of the Product collection type (mainly interesting in typed languages)
• response.paged - whether the query result will be a Page of results