CloudWatch Metrics

💡 Definition

CloudWatch Metrics are data points about your AWS resources published to CloudWatch. A metric represents a time-ordered set of data points that are either published by AWS services or by your own applications.

🔑 Key Concepts

• Time-Series Data: Data points monitored over time (e.g., CPU Utilization every minute).
• Namespaces: Containers for metrics (e.g., AWS/EC2 for EC2 metrics).
• Dimensions: Name/value pairs that identify the metric (e.g., InstanceId=i-12345).
• Custom Metrics: You can publish your own application metrics (e.g., "PageViews", "OrdersProcessed").

⚙️ How it Works

AWS services automatically send metrics to CloudWatch (e.g., EC2 sends CPU, Disk I/O, Network). You can view these in the console or retrieve them via API. You can create graphs and dashboards from them.

🎯 Use Cases

• Performance Monitoring: Tracking if your server is overloaded.
• Operational Health: Monitoring error rates of your API Gateway.
• Billing: Billing data is stored as metrics in the AWS/Billing namespace.

💰 Pricing Model

• Standard Metrics: (Provided by AWS services) are often free or low cost.
• Custom Metrics: Charged per metric per month.
• API Requests: Charged for GetMetricData calls.

📝 Exam Tips (CLF-C02)

• Metrics are the fundamental data unit of CloudWatch.
• Used to trigger CloudWatch Alarms.
• EC2 Memory usage is not a default metric (requires a custom agent).

---

See Also: * CloudWatch * CloudWatch Alarms