A/B Cloud - AWS Cloudfront edge proxy

A Lambda@Edge function used to enable a/b testing, canary releasing, gatekeeping, and SEO a/b/n testing. Enable SSL, for any domain, with AWS certificate manager and Cloudfront. Once set up, abcloud-edge-proxy enables you to a/b test, canary release, and point your domain to any backend, across any cloud provider. Analytics and experimental design (power calculation, statistical analysis) are not part of this package.

Usage

npm install --save abcloud-edge-proxy

Deploy as an Edge Lambda function, within an AWS Cloudfront distribution, for the viewer-request event.

All assignments are done deterministically, by hashing a salt and visitor Id. On every visit the proxy creates a unique request id (using uuid/v4) and forwards this to the origin as a request-id header. On first visits the proxy will use this id as the assignment id. To assure consistent hashing your application MUST set a _vq cookie with this value. On the first request the value of this cookie should be set from the request-id header and on subsequent visits it should come from the _vq cookie (not from the request-id header). For example, in an express app, that might be done as follows:

// set or reset the visitor ID cookie
res.cookie("_vq", req.cookies["_vq"] || req.headers["request-id"], {
    maxAge: 3600 * 1000 * 24 * 365
});

A/B/N Testing Example

import edgeProxy from "abcloud-edge-proxy";

const config = {
    defaultBackend: {
        domainName: "blue.my-site.com"
    },
    origins: [
        {
            domainName: "blue.my-site.com"
        },
        {
            domainName: "green.my-site.com"
        },
        {
            domainName: "red.my-site.com"
        }
    ],
    salt: "unique-assignment-salt",
    test: true
};

exports.handler = edgeProxy(config);

When configuring backends you may use any of the cloudfront parameters, as shown below. Only domainName is required. Current defaults are shown below.

{
    domainName: "required.com",
    port: 443,
    protocol: "https",
    path: "",
    sslProtocols: ["TLSv1", "TLSv1.1"],
    readTimeout: 5,
    keepaliveTimeout: 30,
    customHeaders: {}
};

Canary Releasing Example

Canary releasing can be used to gradually shift traffic from one backend to another. It should ONLY be used with two backends, (unlike a/b/n testing), so that users do not get reassigned as the traffic percentage is increased. An example config is shown below. To assure consistent assignment, for visitors, the weight parameter should only be increased.

import edgeProxy from "edge-proxy";

const config = {
    defaultBackend: {
        domainName: "blue.my-site.com",
    }
    canaryBackend: {
        domainName: "green.my-site.com"
    },
    salt: "unique-assignment-salt",
    canary: true,
    weight: 50 // an integer from 0-100 specifying
    // the percentage of traffic to allocate to the
    // canary backend
};

exports.handler = edgeProxy(config);

Gatekeeping

To enable gatekeeping, you must pass a JWT_SECRET_KEY with the config.

import edgeProxy from "edge-proxy";

const config = {
    JWT_SECRET_KEY: process.env["JWT_SECRET_KEY"]
    /* ... */
};

exports.handler = edgeProxy(config);

You can then encode desired backends into a JWT with your secret and then access your root domain with the JWT set as a query parameter ?devtoken=JWT or as a cookie devtoken=JWT. This allows for ANY backend to be accessed through the proxy and can be useful for developing / testing, on your production domain, without having to update the proxy to add development backends.

An example of creating such a token is shown below.

var jwttoken = require("jsonwebtoken");

const devtoken = jwttoken.sign(
    {
        domainName: "your-desired-backend.com"
        // optional parameters
        // port: 443,
        // protocol: "https",
        // path: "/test",
        // sslProtocols: ["TLSv1", "TLSv1.1"],
        // readTimeout: 60,
        // keepaliveTimeout: 60,
        // customHeaders: {}
    },
    process.env["JWT_SECRET_KEY"]
);

console.log(devtoken);

Search Engine Optimization - A/B Testing Example

Implement by setting SEOTest to true in the config.

Search engine optimization a/b testing is a technique used to validate changes that may impact search rankings. With a/b/n testing, as implemented in this proxy, a unique visitor id is used to hash to a backend. With SEO based a/b/n testing, the full path of each individual request is hashed and used to select the backend. After completion of the experiment period, traffic volumes between the two, or more, SEO implementation are compared for statistical significance.

When running an a/b test, on natural search traffic, it would be wise to validate that your website has enough ranked pages such that a random split of urls results in a generally equal split of natural search traffic (pre-test).

Example config.

import edgeProxy from "abcloud-edge-proxy";

const config = {
    defaultBackend: {
        domainName: "blue.my-site.com"
    },
    origins: [
        {
            domainName: "blue.my-site.com"
        },
        {
            domainName: "green.my-site.com"
        }
    ],
    salt: "unique-assignment-salt",

    // Note the SEOTest flag is set to true
    SEOTest: true
};

exports.handler = edgeProxy(config);

Echoer

To enable the echoer, you must pass an ECHO_TOKEN with the config. You may then access /proxy-echo?echotoken=ECHO_TOKEN and you will get the request object (after modification) returned as formatted JSON. Useful for debugging.

import edgeProxy from "edge-proxy";

const config = {
    ECHO_TOKEN: process.env["ECHO_TOKEN"]
    /* ... */
};

exports.handler = edgeProxy(config);

Building with Webpack

Example building with webpack. Note that edge functions do not currently support environment variables, so we are using webpack to build them into our deployment package.

var path = require("path");
var webpack = require("webpack");

module.exports = {
    entry: {
        proxy: path.join(__dirname, "/index.js")
    },
    output: {
        path: path.join(__dirname, "dist"),
        filename: "edge-proxy-lambda.js",
        libraryTarget: "commonjs"
    },
    target: "node",
    module: {
        rules: [
            {
                test: /\.js$/,
                exclude: /(node_modules)/,
                use: {
                    loader: "babel-loader"
                }
            }
        ]
    },
    plugins: [
        // env vars
        new webpack.DefinePlugin({
            "process.env": {
                JWT_SECRET_KEY: JSON.stringify(process.env["JWT_SECRET_KEY"]),
                ECHO_TOKEN: JSON.stringify(process.env["ECHO_TOKEN"])
            }
        })
    ]
};

After building with webpack zip the function and deploy to AWS.