Posted On: Dec 2, 2022

AWS AppConfig has released an agent for container runtimes that makes the use of feature flags and runtime configuration simpler, with improved performance. Customers that use Amazon Elastic Container Service (Amazon ECS), Amazon Elastic Kubernetes Service (Amazon EKS), Docker, or Kubernetes can now use the AWS AppConfig Agent to manage the calls from their containerized application to the AWS AppConfig service. Previously, customers needed to manage their own retrieval and caching of configuration data when using a container runtime. Now, by using the AWS AppConfig Agent, the agent will do this for them. This agent will query AWS AppConfig for configuration data and make it available locally. It will also handle polling and caching logic on behalf of customers. Applications will make local HTTP calls to the AWS AppConfig Agent, thus seeing performance improvements when retrieving configuration data.

Using feature flags and runtime configuration, customers can change the way an application behaves while running, without deploying new code. For example, when a customer wants to release a new feature, the customer can hide the feature behind a feature flag configuration, and push code to production. Once the feature is ready to be launched, AWS AppConfig will flip the feature flag configuration to 1% of users, then 10%, then 20%, etc. Releasing the feature gradually will help avoid surprises or even outages caused by this change. AWS AppConfig also has additional configuration safety features, including validators to ensure your changes are what you expect, as well as automatic rollback of configuration changes if an Amazon CloudWatch alarm alerts during deployment.

The AWS AppConfig Agent is available in the Amazon Elastic Container Registry (ECR), and can be used in all commercial and AWS GovCloud (US) Regions. To get started, use the AWS AppConfig Getting Started Guide or our AWS AppConfig Agent documentation.