The S3 download speed benchmarks requires a large binary file to be downloaded from a remote S3 repository.
A binary file of an arbitrary size can be created with:
fallocate -l 100M /path/to/file.imgThe source video file is based on the Big Buck Bunny film ((c) copyright 2008, Blender Foundation / www.bigbuckbunny.org) and is available in the video-processing directory.
The model training benchmark reuses the reviews10mb.csv dataset available in the original kmu-bigdata/serverless-faas-workbench repository.
The image processing benchmark reuses the image.jpg file available in the original kmu-bigdata/serverless-faas-workbench repository.
The json benchmark uses the large JSON dataset available in the json-iterator/test-data repository.