Resilience / Scenario lab
Failure, recovery & optimization
Stress the service topology, inspect blast radius, and export a reproducible synthetic research summary.
NTS service topology
Select a component to inspect it
2/4 pods serving
Client pool
UDP/123TCP/4460
Kubernetes cluster
3 WORKERSSelected
nts-pod-2
Estimated capacity
56K qps
Recovery objective
44s
Capacity recovery
Modeled service throughput after scenario injection at t=0
Deployment trade-offs
Comparative assumptions at equivalent CPU allocation
| Platform | Availability | Capacity | Recovery | Ops |
|---|---|---|---|---|
| Bare Metal | 99.90% | Highest | Manual / 15m | |
| Docker | 99.95% | High | Host / 4m | |
| Kubernetes | 99.675% | Degraded | 44s |
Interpretation: orchestration improves automated recovery and placement options, while adding packet-path and operational overhead. Values are scenario assumptions, not production SLO evidence.