Container Technology is becoming the main stream for Software Development & adopting DevOps Practices in an Organization. Let’s get to know why we need to adopt containers technology and What Services are available on AWS to use Containers and How to run containers workload on AWS?
They reason to adopt the containers for your application lifecycle are developers are looking for re platform their whole application with respect to Cost, density better developer parity, agility easy for developer to run locally but the Motivation comes from Brownfield development in which the idea is to modernize the legacy application.
Considerations for running containers workload on AWS
Do they want to run ECS or Kubernetes? Developer and Ops team are capable to manage the Container Enviroment cluster themselves or they need fully managed services provided by AWS? There are the few common question which comes into mind when selecting the container service on AWS.
Managed Containers are mature and out quite of while, Tight integration with IAM, Inside VPC, Cloud Watch logs, metric, Cloud formation. For Example Elastic Container Service (ECS) which is fully managed service to run Docker container is committed to AWS only as this aws specific model
Layer you tell what type of services you want to run and scheduling done by ECS and it run on ec2 or Fargate (Server less Container Platform). Aws managed scheduling and fargateengine is where containers run on.
K8S is short name of Kubernetes and why we choose? K8s is Cloud agnostic not tight to specific cloud provider. Less native integration with cloud provider but on AWS it is available as EKS Managed K8S services. Master node control plane. This is less aws specific at this point of time but the benefit is getting upstream experience of Kubernetes, not hiding anything behind aws sdk, kubectl etc.
Differences between Each Service
ECS vs EKS scheduling layer, control plane, Not actual work run on ECS,EKS just to manage the state of containers. Managing the deployment.
Fargate is not doing scheduling, Still ECS is scheduling and fargate is engine.
EKS is not suppoted currently fargate kind of service.
ECS proprietary to AWS, EKS is Managed Kubernetes Service.
Cost Considerations – containers workload on AWS
ECS is free no cost for scheduling , Ec2 cost will be there. Fargate launch type- lambda model memory CPU over a duration pay per sec, min of 1 minute. Paying premium to use fargate but you are getting is no longer ec2EKS control plan 20 cent per hour,
Cloud Kinetics is always keen to provide the quality services like running containers workload on AWS to its Customer and our team is well versed to help you to start you container world journey. Get in touch with us.
Follow us on LinkedIn for more updates on containers workload on AWS.