Ecs

AWS ECS

Amazon Elastic Container Service (ECS) is a fully managed container orchestration service. Run containers on AWS Fargate (serverless) or EC2 instances.

Table of Contents

• [Core Concepts](#core-concepts)
• [Common Patterns](#common-patterns)
• [CLI Reference](#cli-reference)
• [Best Practices](#best-practices)
• [Troubleshooting](#troubleshooting)
• [References](#references)

Core Concepts

Cluster

Logical grouping of tasks or services. Can contain Fargate tasks, EC2 instances, or both.

Task Definition

Blueprint for your application. Defines containers, resources, networking, and IAM roles.

Task

Running instance of a task definition. Can run standalone or as part of a service.

Service

Maintains desired count of tasks. Handles deployments, load balancing, and auto scaling.

Launch Types

| Type | Description | Use Case | |------|-------------|----------| | Fargate | Serverless, pay per task | Most workloads | | EC2 | Self-managed instances | GPU, Windows, specific requirements |

Common Patterns

Create a Fargate Cluster

AWS CLI:

# Create cluster
aws ecs create-cluster --cluster-name my-cluster

# With capacity providers
aws ecs create-cluster \
  --cluster-name my-cluster \
  --capacity-providers FARGATE FARGATE_SPOT \
  --default-capacity-provider-strategy \
    capacityProvider=FARGATE,weight=1 \
    capacityProvider=FARGATE_SPOT,weight=1

Register Task Definition

cat > task-definition.json  100
aws cloudwatch put-metric-alarm \
  --alarm-name queue-scale-out \
  --metric-name ApproximateNumberOfMessagesVisible \
  --namespace AWS/SQS \
  --dimensions Name=QueueName,Value=my-queue \
  --statistic Average \
  --period 60 \
  --evaluation-periods 1 \
  --threshold 100 \
  --comparison-operator GreaterThanThreshold \
  --alarm-actions 

# Scale-in alarm: queue empty for 3 periods (conservative to avoid flapping)
aws cloudwatch put-metric-alarm \
  --alarm-name queue-scale-in \
  --metric-name ApproximateNumberOfMessagesVisible \
  --namespace AWS/SQS \
  --dimensions Name=QueueName,Value=my-queue \
  --statistic Average \
  --period 60 \
  --evaluation-periods 3 \
  --threshold 0 \
  --comparison-operator LessThanOrEqualToThreshold \
  --alarm-actions 

Fargate Spot interruption handling: Spot tasks receive a SIGTERM 2 minutes before termination. Catch it in your application for graceful shutdown. For SQS consumers, call ChangeMessageVisibility on in-flight messages so they return to the queue rather than timing out.

Auto Scaling

# Register scalable target
aws application-autoscaling register-scalable-target \
  --service-namespace ecs \
  --resource-id service/my-cluster/web-service \
  --scalable-dimension ecs:service:DesiredCount \
  --min-capacity 2 \
  --max-capacity 10

# Target tracking policy
aws application-autoscaling put-scaling-policy \
  --service-namespace ecs \
  --resource-id service/my-cluster/web-service \
  --scalable-dimension ecs:service:DesiredCount \
  --policy-name cpu-target-tracking \
  --policy-type TargetTrackingScaling \
  --target-tracking-scaling-policy-configuration '{
    "TargetValue": 70.0,
    "PredefinedMetricSpecification": {
      "PredefinedMetricType": "ECSServiceAverageCPUUtilization"
    },
    "ScaleOutCooldown": 60,
    "ScaleInCooldown": 120
  }'

CLI Reference

Cluster Management

| Command | Description | |---------|-------------| | aws ecs create-cluster | Create cluster | | aws ecs describe-clusters | Get cluster details | | aws ecs list-clusters | List clusters | | aws ecs delete-cluster | Delete cluster |

Task Definitions

| Command | Description | |---------|-------------| | aws ecs register-task-definition | Create task definition | | aws ecs describe-task-definition | Get task definition | | aws ecs list-task-definitions | List task definitions | | aws ecs deregister-task-definition | Deregister version |

Services

| Command | Description | |---------|-------------| | aws ecs create-service | Create service | | aws ecs update-service | Update service | | aws ecs describe-services | Get service details | | aws ecs delete-service | Delete service |

Tasks

| Command | Description | |---------|-------------| | aws ecs run-task | Run standalone task | | aws ecs stop-task | Stop running task | | aws ecs describe-tasks | Get task details | | aws ecs list-tasks | List tasks |

Best Practices

Security

• Use task roles for AWS API access (not access keys)
• Use execution roles for ECR/Secrets access
• Store secrets in Secrets Manager or Parameter Store
• Use private subnets with NAT gateway
• Enable CloudTrail for API auditing

Performance

• Right-size CPU/memory — monitor and adjust
• Use Fargate Spot for fault-tolerant workloads (70% savings)
• Enable container insights for monitoring
• Use service discovery for internal communication

Reliability

• Deploy across multiple AZs
• Configure health checks properly
• Set appropriate deregistration delay
• Use circuit breaker for deployments

aws ecs update-service \
  --cluster my-cluster \
  --service web-service \
  --deployment-configuration '{
    "deploymentCircuitBreaker": {
      "enable": true,
      "rollback": true
    }
  }'

Cost Optimization

• Use Fargate Spot for batch workloads
• Right-size task resources
• Scale to zero when not needed
• Use capacity providers for mixed Fargate/Spot

Troubleshooting

Task Fails to Start

Check:

# View stopped tasks
aws ecs describe-tasks \
  --cluster my-cluster \
  --tasks $(aws ecs list-tasks --cluster my-cluster --desired-status STOPPED --query 'taskArns[0]' --output text)

Common causes:

• Image not found (ECR permissions)
• Secrets access denied
• Network configuration (subnets, security groups)
• Resource limits exceeded

Container Keeps Restarting

Debug:

# Check CloudWatch logs
aws logs get-log-events \
  --log-group-name /ecs/web-app \
  --log-stream-name "ecs/web/abc123"

# Check task details
aws ecs describe-tasks \
  --cluster my-cluster \
  --tasks task-arn \
  --query 'tasks[0].containers[0].{reason:reason,exitCode:exitCode}'

Causes:

• Health check failing
• Application crashing
• Out of memory

Live Debugging with ECS Exec

Connect directly to a running container without SSH. Requires enableExecuteCommand: true on the service and the SSM agent in your container image (included in most base images).

# Enable on existing service
aws ecs update-service \
  --cluster my-cluster \
  --service web-service \
  --enable-execute-command

# Get a shell in a running task
TASK_ARN=$(aws ecs list-tasks --cluster my-cluster --service-name web-service \
  --query 'taskArns[0]' --output text)

aws ecs execute-command \
  --cluster my-cluster \
  --task $TASK_ARN \
  --container web \
  --interactive \
  --command "/bin/sh"

Requirements: Task role must have ssmmessages:CreateControlChannel, ssmmessages:CreateDataChannel, ssmmessages:OpenControlChannel, ssmmessages:OpenDataChannel permissions.

Service Stuck Deploying

# Check deployment status
aws ecs describe-services \
  --cluster my-cluster \
  --services web-service \
  --query 'services[0].deployments'

# Check events
aws ecs describe-services \
  --cluster my-cluster \
  --services web-service \
  --query 'services[0].events[:5]'

Causes:

• Health check failing on new tasks
• Not enough capacity
• Target group health checks failing

Cannot Pull Image from ECR

Check execution role has:

{
  "Effect": "Allow",
  "Action": [
    "ecr:GetAuthorizationToken",
    "ecr:BatchCheckLayerAvailability",
    "ecr:GetDownloadUrlForLayer",
    "ecr:BatchGetImage"
  ],
  "Resource": "*"
}

Also check:

• VPC endpoint for ECR (if private subnet)
• NAT gateway (if private subnet)
• Security group allows HTTPS outbound

References

• ECS Developer Guide
• ECS API Reference
• ECS CLI Reference
• boto3 ECS

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: itsmostafa
• Source: itsmostafa/aws-agent-skills
• License: MIT

Install and usage instructions live in the source repository linked above.