Skip to content

high-cpu.md

Latest commit

 

History

History
42 lines (32 loc) · 963 Bytes

high-cpu.md

File metadata and controls

42 lines (32 loc) · 963 Bytes
Raw

Runbook: High CPU

Severity: P2 | Team: Infrastructure

Symptoms

  • CPU > 80% sustained for 5+ minutes
  • Grafana alert: HighCPUUsage
  • Slow API responses / timeouts

Diagnosis

# Identify top CPU consumers
kubectl top pods -n gistpin --sort-by=cpu

# Check recent deployments
kubectl rollout history deployment/gistpin-backend

# Inspect pod logs for hot loops
kubectl logs -n gistpin -l app=gistpin-backend --tail=100

Resolution

  1. Scale out if load-driven: 
    kubectl scale deployment/gistpin-backend --replicas=5
  2. Rollback if caused by a bad deploy: 
    kubectl rollout undo deployment/gistpin-backend
  3. Kill runaway pod if isolated: 
    kubectl delete pod <pod-name> -n gistpin

Escalation

If unresolved after 15 min → page on-call engineer via PagerDuty.

Post-Mortem

File a post-mortem within 48 hours using the template in incident-response.md.