Cloud Watch matic Overview

• CloudWatch provides metrics for every services in AWS
• Metric is a variable to monitor (CPUUtilization, Networkln...)
• Metrics have timestamps
• Can create CloudWatch dashboards of metrics

Important Metrics

• EC2 instances: CPU Utilization, Status Checks, Network (not RAM)
  • Default metrics every 5 minutes
  • Option for Detailed Monitoring ($$$): metrics every | minute
• EBS volumes: Disk Read/Writes
• S3 buckets: BucketSizeBytes, NumberOfObjects, AllRequests
• Billing: Total Estimated Charge (only in us-east-I)
• Service Limits: how much you've been using a service API
• Custom metrics: push your own metrics

Summary

• CloudWatch:
  • Metrics: monitor the performance of AWS services and billing metrics
  • Alarms: automate notification, perform EC2 action, notify to SNS based on metric
  • Logs: collect log files from EC2 instances, servers, Lambda functions...
  • Events (or EventBridge): react to events in AWS, or trigger a rule on a schedule
• CloudTrail: audit API calls made within your AWS account
• CloudTrail Insights: automated analysis of your CloudTrail Events
• X-Ray: trace requests made through your distributed applications
• AWS Health Dashboard: status of all AWS services across all regions
• AWS Account Health Dashboard: AWS events that impact your infrastructure
• Amazon CodeGuru: automated code reviews and application performance recommendations