EKS vs ECS vs Fargate: Choosing the Right Container Service

One of the most common questions I get from students and colleagues alike: "We're containerizing our app — should we use EKS, ECS, or Fargate?" Having run production workloads on all three, here is the decision framework I actually use.

The three options in one paragraph each

ECS (Elastic Container Service) is AWS's own orchestrator. It is simple, deeply integrated with IAM, ALB and CloudWatch, and has no control-plane cost. EKS (Elastic Kubernetes Service) is managed Kubernetes — you get the entire Kubernetes ecosystem (Helm, ArgoCD, operators) at the price of more operational complexity and a $73/month control plane. Fargate is not an orchestrator at all: it is a serverless compute engine that runs containers for either ECS or EKS so you never manage EC2 instances.

When I pick ECS

• ▸Small-to-medium teams without existing Kubernetes expertise.
• ▸Workloads that live entirely inside AWS — ECS's IAM integration is the smoothest in the industry.
• ▸When you want the fewest moving parts: task definition, service, ALB, done.

When I pick EKS

• ▸You need the Kubernetes ecosystem: Helm charts, ArgoCD GitOps, custom operators, Istio.
• ▸Multi-cloud or hybrid strategy where portability genuinely matters.
• ▸Your team already knows kubectl — the learning curve is paid off.
• ▸Complex deployment strategies (canary, blue/green with fine control) via Argo Rollouts.

When Fargate makes sense

Fargate removes node management entirely — no AMI patching, no capacity planning, no unused headroom. You pay roughly 20% more per vCPU-hour than equivalent EC2, but for spiky or low-utilization workloads it is usually cheaper overall because you pay only for what tasks actually request. In my experience: batch jobs, internal tools, and services with unpredictable traffic are perfect for Fargate. High-throughput, steady-state services are cheaper on EC2-backed capacity, especially with Savings Plans.

Real numbers from a real workload

For a service needing 2 vCPU / 4 GB continuously in ap-south-1: Fargate ≈ $71/month, on-demand t3.medium-ish EC2 capacity ≈ $45/month, with a 1-year Savings Plan ≈ $28/month. At steady state, EC2 wins. But add the engineering hours spent patching AMIs and right-sizing ASGs, and the gap narrows fast for small teams.

My default recommendation

Start with ECS on Fargate. It is the lowest-friction path to production. Move to EKS only when a concrete requirement demands Kubernetes — not because it looks better on a resume. And if you do go EKS, invest in GitOps (ArgoCD) from day one; retrofitting it later is painful.