ECS Cluster Setup and all details for All Environment

DevCluster Cluster configuration(Infrastructure)

AWS Fargate (Serverless)

Amazon EC2 instances

every thing will taken care by AWS

On-demand

spot

Operating system/Architecture

Amazon Linux 2023

not suitable for any env. because aws will take back that resources any time

EC2 instance type

need to fill

Desired capacity(Min and Max)

need to fill

SSH Key pair

no

Auto Scaling group (ASG)

yes

Monitoring(optional)

Use Container Insights(ON/OFF)

collects metrics for many resources, such as CPU, memory, disk, and network

Tags - optional

QaCluster Cluster configuration(Infrastructure)

AWS Fargate (Serverless)

Amazon EC2 instances

every thing will taken care by AWS

On-demand

spot

Operating system/Architecture

Amazon Linux 2023

not suitable for any env. because aws will take back that resources any time

EC2 instance type

need to fill

Desired capacity(Min and Max)

need to fill

SSH Key pair

no

Auto Scaling group (ASG)

yes

Monitoring(optional)

Use Container Insights(ON/OFF)

collects metrics for many resources, such as CPU, memory, disk, and network

Tags - optional

ProdCluster Cluster configuration(Infrastructure)

AWS Fargate (Serverless)

Amazon EC2 instances

every thing will taken care by AWS

On-demand

spot

Operating system/Architecture

Amazon Linux 2023

not suitable for any env. because aws will take back that resources any time

EC2 instance type

need to fill

Desired capacity(Min and Max)

1

SSH Key pair

yes

Auto Scaling group (ASG)

no

Monitoring(optional)

Use Container Insights(ON/OFF)

collects metrics for many resources, such as CPU, memory, disk, and network

Tags - optional

In AWS ECS Reserved Instances are not available.

Infrastructure requirements for Task definition in ECS

DevCluster

QaCluster

ProdCluster

Launch type

AWS Fargate (Serverless)

0

5

5

Amazon EC2 instances

6

4

5

OS, Architecture

Linux/ARM64

1

0

0

Network mode

Bridge

0

5

5

awsvpc

6

4

5

Task size

CPU

0

5

5

Memory

6

4

5

Task roles - conditional

0

1

3

Task execution role

0

1

3

Container - info

DevCluster

QaCluster

ProdCluster

Resource allocation limits

CPU/GPU

0

5

5

Memory soft limit

6

4

5

Memory hard limit

6

4

5

Environment variables

0

1

3

Log collection

0

1

3

Health Check - optional

0

1

3

Infrastructure requirements for Service in ECS

Service - info

DevCluster

QaCluster

ProdCluster

Compute configuration (advanced)

Capacity provider strategy

0

5

5

launch type (FARGATE/EC2)

6

4

5

Service type

Replica

6

4

5

Daemon

2

4

5

Deployment type

Rolling update

2

4

5

Blue/green deployment(only for AWS codeDeploy)

2

4

5

Deployment failure detection

Rollback on failures: If the current deployment fails, the service is rolled back to the last completed deployment state.

2

4

5

Service Connect (sidecar)

internal communication b/w each microservices (enable/disable)

sidecar log collection (enable/disable)

Load balancer

(enable/disable)

Service auto-scaling - (optional) Scaling policy type

Minimum number of tasks

Maximum number of tasks

2

Target Tracking / Step scaling