At Capital One, we are using Docker and container technologies to advance microservices adoption, increase efficiencies of cloud resources, and decouple the application layer from the underlying infrastructure. Capital One is a federated organization with a “you build it, you own it” culture that provides autonomy and speed for delivery teams. Each federated team runs and operates their container management stack. In order for the federated teams to accelerate their cloud and container-based apps adoption, we created self-service automation tools for creation and operations management of container management stack.
In this session, we explore our push-button automation tool that includes capabilities such as the creation and management of Amazon ECS clusters, an Application Load Balancer for dynamic and context-based routing and provides a user interface via a Jenkins Job or a AWS Lambda function. Our tooling also includes a home-grown dynamic service discovery and routing for applications requiring two-way mutual SSL authentication. We talk through how Capital One regularly updates AMIs with the latest patches and software versions using an automated solution that leverages AWS Lambda to rehydrate the Amazon ECS compute cluster with the latest AMI without causing any downtime. We also discuss how we created a sophisticated canary deployment automation using AWS Lambda and application services, where users can specify how to migrate to a new version of containerized apps and manage the deployment.
AWS empowers enterprise Docker deployment with Amazon ECS and an ecosystem of cloud services and serverless architectures, making containerization in mission-critical environments easier than ever.