§ Automation · Auto-Scaling Policies
Auto-Scaling Policy Manager
Define automatic responses to resource bottlenecks. Trigger workload re-routing or cloud capacity provisioning based on real-time metrics thresholds.
Active Policies
0
Total Policies
0
Lifetime Triggers
0
💡 Traffic Spike Policy Guide
- Trigger: Monitor requests/sec, tokens/sec, or concurrent requests in real-time
- Detection: When spike sustained for N seconds, automatically execute offloading rules
- Offloading: Each rule moves a percentage of traffic from one route (local/mesh) to another (cloud)
- Example: If req/s > 500 for 60s, move 80% of mesh traffic to cloud, auto-recover after 15 min
- Cooldown: Prevents policy from re-triggering immediately after execution