Skip to content

open-feature/playground

Repository files navigation

OpenFeature Logo

Welcome to the OpenFeature playground

The OpenFeature playground is a great place to familiarize yourself with the core concepts and features available in OpenFeature. If you're brand new to feature flagging, consider reviewing the What are feature flags? section in our documentation before running the demo.

Pre-requisites

In order to run the demo, you'll need the following tools available on your system.

How to run the demo

  1. Clone the repo

    git clone https://github.com/open-feature/playground.git
  2. Navigate to the playground folder

    cd playground
  3. Copy .env.example to .env

    cp .env.example .env
  4. Optionally, add feature flag vendors to the .env. See below for more information.

  5. Start the demo

    docker compose up
  6. Open your favorite browser and navigate to http://localhost:30000/

What's in the demo?

The demo consists of three different scenarios where feature flags are used. They help the fictional company Fib3r safely test and release new features. Two of the flags are client-side flags, which are evaluated in the web browser. One is a server-side flag, which is evaluated on the web server.

Rebranding

As we all know, naming is hard! In this scenario, the team at Fib3r is in the process of rebranding from FaaS to Fib3r. This may seem like a situation where a feature flag is unnecessary. However, many times, a rebranding needs to correspond with a press release or blog post. Of course, you could time a deployment moments before the announcement, but that's potentially risky and may require coordination across multiple teams. Using a feature flag would allow you to deploy when it's convenient, test in production by enabling the feature for a subset of users, and then enable it instantly for everyone.

For the rebranding effort, we're only interested in being able to toggle the new welcome message on and off. A boolean value is exactly what we need! That can be accomplished in OpenFeature like this.

Experimenting with color

The team at Fib3r has a hypothesis. They feel that the reason Fib3r hasn't achieved unicorn status is that the current color of the landing page is responsible for high bounce rates. This is a great opportunity to use feature flags for experimentation. With feature flags, it's possible to measure the impact a change has on the metrics that are important to your business.

Diving into the code, you may notice that an after hook has been defined. Hooks are a powerful feature that can be used to extend OpenFeature capabilities. In this case, the code expects a valid CSS hex color value. However, the person configuring the feature flag in a remote feature flag management tool may not be aware of this requirement. That's where a validation hook could be used to ensure only valid CSS values are returned. In this hook, the evaluated value is tested against a regular expression. If it doesn't match, a warning message is logged, and the hook throws an error. OpenFeature will catch the error and return the default value.

Test in production

Fib3r is on a mission to help the world calculate the nth digit of Fibonacci more efficiently. According to a Stack Overflow article the team recently found, it's possible to use Binet's Formula to calculate Fibonacci more efficiently. While the initial tests look promising, changing the underlying algorithm Fib3r has used for years is risky. The team decided that it would be safer put the new feature behind a context-dependant feature flag so that only employees could use it initially. If the test goes well, the feature could be slowly rolled out to everyone or quickly reverted if an issue is discovered.

Let's see how this could be done using OpenFeature. Here is where the Fib3r team adds a feature flag that returns the name of the algorithm to run. Looking closely at the getStringValue method, you'll notice evaluation context is not being defined. Evaluation context is commonly used in feature flagging to determine the flag value dynamically. For example, the Fib3r team may want to test the binet algorithm on employees only. This can be done by setting the user's email address as evaluation context and defining a rule that returns binet only when the email address ends with @faas.com. Simple enough, but remember that the evaluation context wasn't explicitly set during the flag evaluation linked above. That's because OpenFeature allows developers to set evaluation context at various points in their application. In this case, evaluation context is set on each transaction and automatically used during flag evaluation.

Available providers

The following providers can be used in the demo. Locate the provider you're interested in using to learn more.

Environment Variable

The environment variable provider is a simple demo showing how environment variables could be used to make flag evaluations. Its purpose is to show how a basic provider could be implemented.

To get started, follow the instructions in the How to run the demo section. Once in the demo app, select env from the dropdown located at the bottom-right of the screen. To change a flag value, open the .env file in your favorite text editor. Update the flag values based on the options defined as comments above the flag key. When you're ready, save the file and restart the demo.

Using environment variables like this can be a good way to get started with feature flagging. However, the approach only supports basic use cases and is quite cumbersome.

flagd

flagd is an OpenFeature compliant flag evaluation daemon. Following the Unix philosophy, it provides one component of a full feature flagging solution: a service for storing and evaluating flags. It supports the ability to define flag configurations in various locations, including a local file, an HTTP service, or in the case you're using Kubernetes, directly from the Kubernetes API.

In this demo, FlagD starts automatically as part of the Docker Compose file. It's configured to watch a local file /config/flagd/flags.json for flag configurations. Feel free to modify this file and see how it affects the demo. Valid configuration changes should be reflected almost immediately.

Flag Configuration in FlagD

A FlagD configuration is represented as a JSON object. Feature flag configurations can be found under flags, and each item within flags represents a flag key (the unique identifier for a flag) and its corresponding configuration.

Valid flag configuration options include:

State

state is required property. Validate states are "ENABLED" or "DISABLED". When the state is set to "DISABLED", FlagD will behave like the flag doesn't exist.

Example:

"state": "ENABLED"
Variants

variants is a required property. It is an object containing the possible variations supported by the flag. All the object's values must be the same type (e.g. boolean, numbers, string, JSON). The type used as the variant value will directly affect how the flag is accessed in OpenFeature. For example, to use a flag configured with boolean values, the getBooleanValue or getBooleanDetails methods should be used. If another method, such as getStringValue is called, a type mismatch occurs and the default value is returned.

Example:

"variants": {
  "red": "c05543",
  "green": "2f5230",
  "blue": "0d507b"
}

Example:

"variants": {
  "on": true,
  "off": false
}

Example of an invalid configuration:

"variants": {
  "on": true,
  "off": "false"
}
Default Variant

defaultVariant is a required property. The value must match the name of one of the variants defined above. The default variant is always used unless a targeting rule explicitly overrides it.

Example:

"variants": {
  "on": true,
  "off": false
},
"defaultVariant": "off"

Example:

"variants": {
  "red": "c05543",
  "green": "2f5230",
  "blue": "0d507b"
},
"defaultVariant": "red"

Example of an invalid configuration:

"variants": {
  "red": "c05543",
  "green": "2f5230",
  "blue": "0d507b"
},
"defaultVariant": "purple"
Targeting Rules

targeting is an optional property. A targeting rule must be valid JSON. FlagD uses a modified version of JSON Logic, as well as some custom pre-processing, to evaluate these rules. The output of the targeting rule must match the name of one of the variants defined above. If an invalid or null value is returned by the targeting rule, the defaultVariant value is used.

The JSON Logic playground is a great way to experiment with new targeting rules. The following example shows how a rule could be configured to return binet when the email (which comes from the evaluation context) contains @faas.com. If the email wasn't included in the evaluation context or didn't contain @faas.com, null is returned, and the defaultVariant is used instead. Let's see how this targeting rule would look in the JSON Logic playground.

  1. Open the JSON Logic playground in your favorite browser
  2. Add the following JSON as the Rule:
{
  "if": [
    {
      "in": [
        "@faas.com",
        {
          "var": ["email"]
        }
      ]
    },
    "binet",
    null
  ]
}
  1. Add the following JSON as the Data:
{
  "email": "test@faas.com"
}
  1. Click Compute
  2. confirm the output show "binet"
  3. Optionally, experiment with different rules and data

Go Feature Flag

Go Feature Flag is a open source feature flagging solution. It can define flag configurations in various locations (HTTP, S3, GitHub, file, Google Cloud Storage, Kubernetes). OpenFeature is able to integrate with Go Feature Flag by using the Go Feature Flag Relay Proxy.

In this demo, Go Feature Flag starts automatically as part of the Docker Compose file. It's configured to watch a local file /config/go-feature-flag/flags.yaml for flag configurations. Feel free to modify this file and see how it affects the demo. Valid configuration changes should be reflected almost immediately. Documentation on how to configure a flag can be found here.

CloudBees Feature Management

CloudBees Feature Management is an advanced feature flagging solution that lets your development teams quickly build and deploy applications without compromising on safety. By providing a gradual release mechanism and a simple way to define target audiences, CloudBees Feature Management allows developers and product managers to optimize feature releases and customize the user experience. CloudBees Feature Management gives teams control over features that are in staging, production, or any environment in the deployment pipeline.

Follow these steps to set up CloudBees for the demo:
  1. Sign in to your CloudBees Feature Management account. If you don't already have an account, you can use the free community edition.

  2. Within the CloudBees Feature Management UI, add a new application called OpenFeature playground. You can keep the default environment of production.

  3. In App Settings add a new custom STRING property called email as shown below. This is used in the fib-algo configuration to control the flag value via the email of the user logging into the playground application.

  4. Create a new boolean flag called new-welcome-message.

  5. Create a new flag called hex-color with the values: c05543, 2f5230, and 0d507b.

  6. Create a new flag called fib-algo with the values: recursive, memo, loop, binet, and default.

  7. For the fib-algo flag, add a configuration. This can be a combination of the email regEx of .*faas.com$ and set recursive in the else section.

  8. Ensure for each flag, the configuration switch is set to ON, as shown below

  9. Ensure the completed list of flags looks as follows

  10. Copy the production environment key found under App settings > Environments

  11. Open the .env file and make the value of CLOUDBEES_APP_KEY the key copied above

Now that everything is configured, you should be able to start the demo. Once it's started, select cloudbees from the provider list located at the bottom right of your screen. You should now be able to control the demo app via CloudBees! Note that for "UI" flags (hex-color, new-welcome-message) you have to select Platform: Browser in the platform dropdown when modifying flag values.

Split

Split is a unified feature flagging and experimentation platform enabling product and engineering teams to reduce cycle times, mitigate release risk, and maximize business impact throughout the Feature Delivery Lifecycle.

Follow these steps to set up Split for the demo:
  1. Sign in to your Split account. If you don't already have an account, you can use the Split Free Edition.

  2. Create a new split called new-welcome-message and use the default treatments.

  3. Create a new split called hex-color with the treatments: c05543, 2f5230, and 0d507b.

  4. Create a split called fib-algo with the values: recursive, memo, loop, binet, and default.

  5. For the fib-algo flag, add a targeting rule that serves binet if the emails email address ends with @faas.com.

  6. Create a new server-side and client side API keys. This can be done by navigating to Admin settings > API keys > Create API key.

  7. Open the .env file and set the values of SPLIT_KEY and SPLIT_KEY_WEB to the keys copied above.

Now that everything is configured, you should be able to start the demo. Once it's started, select split from the provider list located at the bottom right of your screen. You should now be able to control the demo app via Split!

Harness

Harness Feature Flags is a new approach to feature management that empowers teams to move faster and have more control at the same time. Templatize feature rollouts, automate release workflows, build using GitOps and config as code, and get complete control over what gets released and when, without sacrificing velocity.

Follow these steps to set up Harness for the demo:
  1. Sign in to your Harness account. If you don't already have an account, you can use the free plan.

  2. Use an existing organization and project or create a new one.

  3. Create a new boolean feature flag called new-welcome-message and confirm the ID is newwelcomemessage.

  4. Create a new multivariate feature flag called hex-color and confirm the ID is hexcolor. Add three variants with the following names and values: red - c05543, green - 2f5230, and blue - 0d507b.

  5. Create a new multivariate feature flag called fib-algo and confirm the ID is fibalgo. Add these variants to both the name and value: recursive, memo, loop, binet, and default.

  6. Add a new targeting group called Fib3r Employees by going to Target Management > Target Groups > New Target Group.

  7. Add a targeting condition that looks for the Identifier to end with @faas.com.

  8. Add the fib-algo flag and set the variation to binet.

  9. Create new server and client SDK keys. This can be done by navigating to Environments > Development > New SDK Key.

  10. Open the .env file and set the values of HARNESS_KEY and HARNESS_KEY_WEB to the keys copied above.

Now that everything is configured, you should be able to start the demo. Once it's started, select harness from the provider list located at the bottom right of your screen. You should now be able to control the demo app via Harness!

LaunchDarkly

LaunchDarkly unleashes developer productivity for the software-powered world by fundamentally changing how you deliver software to your customers. With LaunchDarkly's feature management platform, empowered developers can empower the business to release new features faster and more efficiently than ever.

Follow these steps to set up LaunchDarkly for the demo:
  1. Sign in to your LaunchDarkly account. If you don't already have an account, you can sign up for a free trial.

  2. Create a new feature flag with the key new-welcome-message using the default boolean flag variation.

  3. Go to "Settings" for this flag and ensure that "SDKs using Client-side ID" is checked under "Client-side SDK availability".

  4. Create a new feature flag with the key hex-color. Set the flag variations to string and add three variations with the following variation and name: c05543 - red, 2f5230 - green, and 0d507b - blue.

  5. Go to the "Settings" for this flag and ensure that "SDKs using Client-side ID" is checked under "Client-side SDK availability".

  6. Create a new feature flag with the key fib-algo. Set the flag variations to string and add these variants to both the variation and name: recursive, memo, loop, binet, and default.

  7. Select the fib-algo flag and add a targeting rule that looks for the email to end with @faas.com and serves binet.

  8. Navigate to Account settings > Environments and copy the SDK Key and Client-side ID associated with the environment you would like to use.

  9. Open the .env file and set the values of LD_KEY and LD_KEY_WEB to the SDK Key and Client-side ID of the key copied above.

Now that everything is configured, you should be able to start the demo. Once it's started, select launchdarkly from the provider list located at the bottom right of your screen. You should now be able to control the demo app via LaunchDarkly!

Flagsmith

Flagsmith is an open source, fully featured, Feature Flag and Remote Config service. They support a low latency hosted Edge API, deployment to your own private cloud, or running on-premise. Flagsmith makes it easy to create and manage features across web, mobile, and server side applications.

Follow these steps to set up Flagsmith for the demo:
  1. Sign in to your Flagsmith account. If you don't already have an account, you can sign up for the free plan.

  2. Navigate to Environment > Development > Features.

  3. Create a new feature with the id new-welcome-message. Enabled the feature and add a true and false variation values. Confirm that true is the control value and set it to 100%.

  4. Create a new feature with the id hex-color. Enabled the feature and add a c05543, 2f5230, and 0d507b variation values. Confirm that c05543 is the control value and set it to 100%.

  5. Create a new feature with the id fib-algo. Enabled the feature and add a recursive, memo, loop, binet, and default variation values. Confirm that recursive is the control value and set it to 100%.

  6. Create a new segment called fib3r_employees and add a rule that checks if email trait email contains @faas.com.

  7. Select the fib-algo feature under the development environment and add a segment override. Select the fib3r_employees override, enable the override, and confirm the control is set to binet.

  8. Navigate to Environments > Development > Settings and create a new server-side environment key.

  9. Open the .env file and set the values of FLAGSMITH_ENV_KEY the key copied above, and FLAGSMITH_ENV_KEY_WEB to the value of the "Client-side Environment Key" shown in the Flagsmith UI.

Now that everything is configured, you should be able to start the demo. Once it's started, select flagsmith from the provider list located at the bottom right of your screen. You should now be able to control the demo app via Flagsmith!

Flipt

Flipt is an open-source feature management platform that's fully self-hosted. It's easy to set up, has no seat limits, and is built for developers from scale-ups to enterprises.

After starting the demo, the Flipt UI is available at http://localhost:8080.

Experimenting beyond the demo

Evaluation context

The following evaluation context is available during flag evaluation. That means any of these properties can be used when defining a rule in the feature flag manager of your choosing.

Name Description
ip The IP address sent from the browser. This comes from the x-forwarded-for header.
email The email address of the logged-in user. Returns anonymous if the user is logged out
targetingKey Same as email
method The HTTP method used (e.g. GET, POST, PUT)
path The HTTP path of the request (e.g. /hex-color)
ts The current time in milliseconds

Troubleshooting

Ports are not available

Confirm that the following ports are available 30000, 8013, and 16686.

Vendor isn't listed in the dropdown

To add a vendor to the demo, follow the vendor-specific section in the documentation. An SDK key must be added to the appropriate property in the .env file and the demo needs to be restarted.

The UI is always grey

This means that the provider you've select either doesn't support client-side (see what's in the demo section) or it's not working properly. Ensure you've correctly configured the respective client-side provider, including the credentials.