ʕ •́؈•̀) Flash - Response Lantency Simulator

A Cloudflare Worker to simulator response lantency. Use it to test benchmark tool like ab or simulator API service lantency.

🧪 Usage

Currently only GET request is supported. There're servel params to control how the simulator works:

• lmin: Min lantency in ms to add to the response. Default: 10
• lmax: Max lantency in ms to add to the response. Default: 20
• echo: Response body if provided
• proxy: Valid url to fetch and return to client. There's a whitelist to control which hosts can be proxied. The default whitelist is ['www.google.com', 'postb.in']. Deploy your own worker if you want to proxy to more hosts.
• d: Distribution method: uniform | normal | erlang. Default: uniform. For normal distribution, there're two optional params d_mu(mean, default to (lmin + lmax) / 2) and d_sigma(standard deviation, default to 1). For erlang distribution, d_k(shape, default to 1) and d_mu(scale, default to (lmax - lmin) / 2).

For example, curl "http://flash.codeb2cc.com/?lmin=100&lmax=200" will respone in a random delay between [100, 200)ms. curl "http://flash.codeb2cc.com/?proxy=https%3A//www.google.com/" will fetch 'https://www.google.com/' and return in a [10, 20)ms delay. More:

curl "http://flash.codeb2cc.com/?lmin=100&lmax=200&d=normal&d_sigma=0.5"

curl "http://flash.codeb2cc.com/?lmin=100&lmax=200&d=normal&d_mu=130"

curl "http://flash.codeb2cc.com/?lmin=0&lmax=100&d=erlang&d_k=2&d_mu=30"

👩 💻 Developing

You can use this project as a template to make your own Cloudflare Worker:

wrangler generate my-app https://github.com/codeb2cc/cf-flash-slothmore

👀 TODO

More features:

• Different lantency distribution models (normal/erlang(poisson) and more)
• Support simulator params(lmin, lmax, etc) in request headers
• Cache proxy response to protect target API service
• Custom cache missing rate and ttl
• Predefine response data read from Cloudflare KV