Contents:

Overview

Monitoring and logging are essential for tracking the health and performance of your application and infrastructure.
In this workshop, we will use:

  • Amazon CloudWatch Logs to store and analyze logs from containers.
  • CloudWatch Container Insights to monitor CPU, Memory, Network, and Storage usage.
  • CloudWatch Alarms to receive notifications when issues occur.

Enable Amazon CloudWatch Logs

  1. Go to ECS Console → Cluster → Select your service.
  2. In the Task Definition, check the log driver configuration:
    • Select awslogs as the log driver.
    • Log Group name: /ecs/fargate-workshop
    • Region: us-east-1 (or your chosen AWS region).
  3. Save and redeploy the service. Amazon CloudWatch Logs

Enable Container Insights

  1. Go to CloudWatch Console → Container Insights.
  2. Select Enable for your ECS Cluster.
  3. Container Insights will automatically collect:
    • CPUUtilization
    • MemoryUtilization
    • NetworkRxBytes / NetworkTxBytes
    • StorageReadBytes / StorageWriteBytes

Viewing Logs and Metrics

  1. Open CloudWatch Console → Logs → Select the Log Group /ecs/fargate-workshop.
  2. View container logs (stdout, stderr).
  3. Go to CloudWatch Metrics → ECS → Cluster Metrics to see CPU, RAM, and Network statistics.
  4. Filter by specific services or tasks for detailed analysis.

Integrating Alerts (Alarms)

  1. Go to CloudWatch Console → Alarms → Create alarm.
  2. Select a metric, for example: ECS/ContainerInsights – CPUUtilization.
  3. Set a threshold, e.g., CPU > 80% for 5 minutes.
  4. Choose an SNS Topic to receive email notifications when the alarm is triggered.
  5. Save and activate the alarm.

Expected Outcome:

  • You can view logs for each container to debug issues.
  • Monitor application performance in real-time.
  • Receive notifications when the system encounters issues or exceeds resource thresholds.