{
  "serviceArguments": {
    "externalIP": "192.0.2.10",
    "host": "192.0.2.10",
    "port": 28182,
    "protocol": "tcp",
    "sel": 0,
    "mode": 4,
    "model_name": "Qwen/Qwen2.5-7B-Instruct",
    "sse_mode": true,
    "pd_disagg_mode": true,
    "pd_cache_aware_mode": false,
    "kvExactMode": 0,
    "kvEngineType": "vllm",
    "monitor": true,
    "probetype": "http",
    "probeport": 8000,
    "probereq": "/health",
    "probeTimeout": 1,
    "probeRetries": 1,
    "pd_prefill_timeout_sec": 15,
    "cb_enable": true
  },
  "endpoints": [
    {
      "endpointIP": "192.0.2.20",
      "targetPort": 8000,
      "weight": 1,
      "ep_role": 1,
      "nixl_port": 5600
    },
    {
      "endpointIP": "192.0.2.21",
      "targetPort": 8000,
      "weight": 1,
      "ep_role": 1,
      "nixl_port": 5600
    },
    {
      "endpointIP": "192.0.2.30",
      "targetPort": 8000,
      "weight": 1,
      "ep_role": 2,
      "nixl_port": 5600
    },
    {
      "endpointIP": "192.0.2.31",
      "targetPort": 8000,
      "weight": 1,
      "ep_role": 2,
      "nixl_port": 5600
    }
  ]
}
