Amazon Elastic Kubernetes Service

Amazon Elastic Kubernetes Service
Name	Amazon Elastic Kubernetes Service
Developer	Amazon Web Services
Released	2018
Operating system	Cross-platform
Genre	Container orchestration
License	Proprietary

Amazon Elastic Kubernetes Service. It is a managed container orchestration service that simplifies the deployment, management, and scaling of containerized applications using Kubernetes. The service automates critical tasks such as node provisioning, patching, and updates, allowing teams to focus on application development. As a core offering within the Amazon Web Services ecosystem, it provides a highly available and secure platform for running microservices and other modern application architectures.

Overview

Launched in 2018, this service is a fully managed implementation of the open-source Kubernetes system, originally designed by Google. It enables organizations to run Kubernetes on Amazon Web Services without needing to install and operate their own Kubernetes control plane. The service is certified as Kubernetes conformant by the Cloud Native Computing Foundation, ensuring compatibility with a vast ecosystem of third-party tools and applications. It is widely adopted by enterprises, including companies like Intuit and GoDaddy, to build portable, scalable applications.

Architecture

The service architecture separates the responsibilities of the managed control plane, run by Amazon Web Services, and the worker nodes, which customers manage. The control plane, comprising components like the API server and etcd datastore, is automatically scaled and distributed across multiple Availability Zones for high availability. Worker nodes run on Amazon EC2 instances or AWS Fargate, with the kubelet agent facilitating communication. Networking is typically handled through the Amazon VPC Container Network Interface, and storage is integrated with services like Amazon EBS and Amazon EFS.

Features

Key features include automated version updates and patching for the Kubernetes control plane, integrated monitoring with Amazon CloudWatch, and security compliance with standards like SOC and ISO/IEC 27001. It offers managed node groups for simplified node lifecycle management and supports both x86-64 and Arm-based Graviton processors. For security, it integrates with AWS Identity and Access Management for authentication and Amazon ECR for storing container images. The service also provides tools for troubleshooting and diagnostics through the AWS Management Console and AWS CLI.

Integration with AWS services

The service is deeply integrated with the broader Amazon Web Services portfolio. For compute, it seamlessly works with Amazon EC2 and serverless AWS Fargate. Storage options include persistent volumes from Amazon EBS, Amazon EFS, and Amazon FSx for Lustre. Networking capabilities are enhanced through Amazon VPC, Elastic Load Balancing, and AWS App Mesh. It connects to managed databases like Amazon RDS and Amazon DynamoDB, and leverages AWS Secrets Manager for credential management. Observability is achieved via Amazon CloudWatch, AWS X-Ray, and Amazon Managed Service for Prometheus.

Pricing

Pricing follows a model where users pay only for the Amazon Web Services resources created to run their Kubernetes worker nodes, such as Amazon EC2 instances or AWS Fargate tasks. There is an additional hourly charge for each Kubernetes cluster to cover the managed control plane, with the rate varying by AWS Region. Costs for associated services like Amazon EBS storage, Elastic Load Balancing load balancers, and data transfer fees are billed separately. The AWS Free Tier includes limited usage for new customers to evaluate the service.

Use cases

Common use cases include deploying scalable microservices architectures for web applications, as utilized by companies like Netflix and Snap Inc.. It is instrumental in machine learning workflows, integrating with Amazon SageMaker for model training and deployment. The service supports hybrid cloud deployments and application migration projects, often in conjunction with AWS Outposts. It is also used for running batch processing jobs, continuous integration and delivery pipelines with tools like Jenkins, and stateful applications like Apache Kafka and Elasticsearch.

Category:Amazon Web Services Category:Cloud computing Category:Container orchestration