NIP: host CPU high

Alert: NipHostCPUHigh · Severity: Warning

What this means

Average CPU on the NIP app server exceeded 80% for 5+ minutes (node_exporter).

Business impact

• Sustained high CPU can slow transfers and increase timeout risk.
• May precede worker wedging if all cores saturated.

Confirm

ssh baasey  # or easypay
top -bn1 | head -20
docker stats --no-stream | sort -k3 -hr | head -10

Prometheus: 100 - avg(rate(node_cpu_seconds_total{mode="idle"}[5m]))*100

Fix

1. Identify top container: docker stats.
2. If web dominates — check for traffic spike vs stuck workers (see health probe).
3. If postgres — check slow queries runbook.
4. If host-wide — consider scaling or killing runaway process (careful on prod).

Escalate if

CPU >90% sustained 10+ min or transfers failing.