Feature Flag backend

This app should be deployed in an cluster which is running the FeatureFlag Operator.

Feature flag resources can be deployed in the cluster and their status can be queried using this app.

Usage

This app expose a variety of endpoints to query the status of the feature flags:

The following endpoint can be used to manage the K8s resources itself:

• GET /api/v1/featureflags: List all the feature flags in the cluster. It will return a list of FeatureFlag resources.
• GET /api/v1/featureflags/{name}: Get the status of a specific feature flag. It will return a FeatureFlag resource.
• POST /api/v1/featureflags: Create a new feature flag. DTO is available here
• PUT /api/v1/featureflags: Update the fiven featureflag. DTO is available here
• DELETE /api/v1/featureflags/{name}: Delete the given feature flag.

The following endpoints can be used to query the status of the feature flags and should be used by the applications:

• GET /api/v1/featureflags/{name}/active: return a json with the status of the given FF and if it is enabled or not
• PUT /api/v1/featureflags/{name}/enable: enable a given FF
• PUT /api/v1/featureflags/{name}/disable: disable the given FF
• PUT /api/v1/featureflags/{name}/activate: activate the given FF
• PUT /api/v1/featureflags/{name}/deactivate: deactivate the given FF

All the above endpoints accept the following query parameters:

• namespace: The namespace where the feature flag is deployed. If not provided, it will use the default namespace defined in config.json file.

Glossary

A feature flag can be enabled or not and even active or not:

• Enabled: The feature flag is enabled in the cluster. It means that the feature flag is deployed in the cluster and it is ready to be used. If this value is false, it's status will always be false.
• Active: The feature flag is active. It means that the feature flag is enabled and it is ready to be used. This value can be used by the applications to decide what to show to the users.

Development

Prerequisites

• Go
• Docker
• Kubectl
• Running kubernetes cluster (e.g. minikube)
• FeatureFlag Operator and crds installed in the cluster

Run locally

Run docker-compose up to run the application locally. You must have a k8s cluster running locally and need to copy the kubeconfig file to the root of the project.

Build

This app is using the k8s code generator to build the clientset based on the given CRDs.

You need to install the code-generator tool:

go get k8s.io/code-generator

Rebuild the clientset only if the CRDs are changed:

./hack/update-codegen.sh