HPA baseline misconfiguration causes pod thrashing

HPA configured without proper behavior controls, CPU targets, or min/max replica bounds causes pod thrashing and instability during traffic spikes

Configure HPA v2 with: CPU target 60-70% (65% typical) for burst headroom, min replicas ≥3 for HA, verified max cluster can handle. Set scaleUp to 100% per 60s (stabilizationWindowSeconds: 0), scaleDown to 20% per 60s with 300s stabilization