CloudWatch Alarms

💡 Definition

CloudWatch Alarms allow you to watch a single CloudWatch metric over a time period and perform one or more actions based on the value of the metric relative to a threshold you define.

🔑 Key Concepts

• Threshold: The limit you set (e.g., CPU > 80%).
• States:
  • OK: Within threshold.
  • ALARM: Breached threshold.
  • INSUFFICIENT_DATA: Not enough data to decide.
• Actions: What happens when the alarm changes state.
  • Notification: Send an email/SMS via SNS.
  • Auto Scaling: Trigger a scale-out or scale-in policy.
  • EC2 Action: Stop, terminate, or recover an instance.

⚙️ How it Works

1. Select Metric: Choose CPUUtilization for an instance.
2. Define Condition: "Greater than 80% for 5 minutes".
3. Configure Action: "Send email to Admin" AND "Add 1 instance to ASG".

🎯 Use Cases

• Alerting: Getting notified when your website is slow or down.
• Automation: Automatically scaling your infrastructure to meet demand.
• Remediation: Rebooting a stuck instance.

💰 Pricing Model

• Standard Resolution: Free tier includes 10 alarms. Paid per alarm per month thereafter.
• High Resolution: (10-second intervals) costs more.

📝 Exam Tips (CLF-C02)

• Alarms trigger actions based on metrics.
• Key integration with Auto Scaling and SNS (Simple Notification Service).

---

See Also: * CloudWatch * Auto Scaling * SNS