Back
Monitoring & Alerts
Detect incidents early with SLOs and alert policies.
SLOs
Define availability and latency targets per critical path (e.g., checkout p95 < 300ms, error rate < 0.5%).
Alert policy
- Page only on customer impact; everything else as tickets.
- Use multi-window burn rates for SLO alerts.