Skip to content

Performance Testing

Brian Warner edited this page Jun 20, 2023 · 5 revisions

(preliminary)

Running yarn bench in a package directory will run any existing benchmarks. The results will be printed to stdout.

Benchmarks are stored in files named bench-*.js, just like tests are in test-*/js. The benchmark runner finds all files in bench/**/bench-*.js.

The first benchmarks written are in packages/vats/.

Running yarn bench --save=foo.json will save the results to the named file, in a machine-parsable format.

A CI job will run the benchmarks periodically (nightly, and after each push to trunk), and upload the results to a DataDog dashboard at $URL.

Clone this wiki locally