What it is: A rule that watches a metric and takes action when it crosses a threshold.

What it’s for:

• Notify (SNS), scale (ASG), or trigger automation.
• Detect outages or performance degradation early.

Key ideas:

• Alarm states: OK, ALARM, INSUFFICIENT_DATA.
• Alarms can be used for both ops alerts and auto scaling.

Exam cues:

• “send notification when error rate is high” → alarm + SNS.
• “trigger scale-out when CPU high” → alarm + ASG policy.

Hard words:

• *notify* /ˈnoʊtɪfaɪ/: thông báo
• *degradation* /ˌdeɡrəˈdeɪʃn/: suy giảm
• *insufficient* /ˌɪnsəˈfɪʃnt/: không đủ