Horizontal Pod Autoscaling

Auto-scale Raphtory based on load

Configure HPA to automatically scale Raphtory pods based on CPU, memory, or custom metrics.

Basic HPA Configuration

Apply:

Custom Metrics (Graph Size)

Scale based on graph size or queue depth:

Monitor HPA

Best Practices

1. Min replicas: Set to 2+ for HA
2. Stabilization: Prevent flapping with 5min cooldown
3. Multiple metrics: Combine CPU + memory + custom
4. Test: Load test to verify scaling behavior

See Also

• Resource Limits - Sizing per pod
• Performance Optimization