Kubernetes Autoscaler

Custom Kubernetes HPA with GPU utilization metrics.

Features

• GPU utilization-based scaling (not just CPU/memory)
• Custom Prometheus metrics support
• Predictive scaling with time-series forecasting
• Scale-to-zero for batch workloads

Metrics

metrics:
  - type: Pods
    pods:
      metric:
        name: gpu_utilization
      target:
        type: AverageValue
        averageValue: "70"

License

MIT