Alerting
一句话解释
Alerting 是练习的 automatically notifying team 当 metrics cross thresholds or 错误 occur.
用大白话说
Alerting is how you find out about problems before your users do. You define rules — e.g., "如果错误速率 > 5% for 5 minutes, alert team" — and when the condition is met, the system sends a notification.
Good alerting is a balance. Too few alerts and you miss critical issues. Too many alerts and you get alert fatigue — the team ignores them all. The best alerts are actionable: when an alert fires, there is a clear action to take.
Alerts are routed through an Alertmanager (like Prometheus Alertmanager) to the right destination — Slack for warnings, PagerDuty for critical issues at 3 AM, email for informational alerts. Alerts should include enough context (links to dashboards, logs) to help diagnose the problem.
示例
# Alertmanager routing
route:
receiver: 'slack'
routes:
- match:
severity: critical
receiver: 'pagerduty'
receivers:
- name: 'slack'
slack_configs:
- api_url: 'https://hooks.slack.com/...'
- name: 'pagerduty'
pagerduty_configs:
- service_key: 'your-key'这个 routes 关键 alerts 到 PagerDuty and 所有 others 到 Slack.
工作原理
- You define alert rules (e.g., error rate > 5% for 5 min).
- The monitoring system evaluates rules continuously.
- When a rule fires, it sends an alert to the Alertmanager.
- The Alertmanager routes the alert to the right channel (Slack, PagerDuty, email).
常见混淆
容易混淆的对象: Monitoring
关键区别: Monitoring collects and displays metrics; alerting 是操作 taken 当 metrics indicate problem. Alerting 是部分的 monitoring.
相关术语
解读错误信息
用大白话解释常见错误——什么意思、为什么发生、如何修复。