When the status of certain products changes, you can create alarms to promptly notify you to take action. Alarms monitor specific metrics over a certain period and send alarms at configured intervals based on configured thresholds.
An alarm consists of the following components:
- Alarm trigger condition (under what conditions will the alarm be sent)
- Alarm object (which object will send an alarm)
- Alarm recipient group (who receives the alarm).
- Alarm receiving method (how the alarm is received)
This document describes how to create alarms for one or more objects, and select the objects to receive alarms.
- Basic Concepts
| Term | Definition |
|---|---|
| Alarm policy type | Policy type identifies policy category and corresponds to specific cloud products, with a distinct layer of granularity under cloud products. For example, when you choose Cloud Virtual Machine > Basic Monitoring policy, you can customize alarm metrics such as CPU utilization and disk utilization; when you choose Cloud Virtual Machine > Storage Monitoring policy, you can customize alarm metrics such as disk I/O wait time and disk read traffic. |
| Alarm rule | Alarm rules correspond to the trigger conditions that specific monitoring metrics should satisfy. For example, when disk utilization exceeds 80% per minute and lasts for 3 minutes, an alarm is triggered once every hour. |
| Relationship between policy types and alarm rules | An alarm type contains multiple alarm rules and is a collection of a series of alarm rules. For example, when you choose Cloud Virtual Machine > Basic Monitoring policy, the policy includes disk utilization alarms, ping unreachable alarms, CPU utilization alarms, and more. |
- Alarm statuses
| Alarm Status | Description |
|---|---|
| Unresolved | Alarms that have not been handled or are being handled, but the resources are still in an abnormal status. |
| Resolved | Alarms have been resolved. |
| Insufficient Data | - The alarm policy that triggered an alarm has been deleted. - The CVM has been migrated from one project to another. - No data is reported because the Cloud Monitor control has not been installed or has been uninstalled. |
| Expired | - The alarm policy has changed. - The most recent alarm time has not been updated for more than 24 hours. |