Skip to content
ACS Infra
Health

Health

GET
/health

Return the state of the API in v1's shape: ok, degraded with its degraded_reasons, or down. It checks the database on its own connection, whether a pooled connection is free, and each runner, and never asks Inference. backends gives each served model's breaker and requests in flight, and degraded_backends the models whose breaker is open. 503 only when the database does not answer within HEALTH_TIMEOUT; a degraded API answers 200.

Response Body

Health

application/json
  1. response
status*string
Value in"ok""degraded""down"
wrapper*
backends*array<>
degraded_backends*array<string>
degraded_reasons*array<string>
Example Request
curl -X GET 'https://api.v2.infra.acsresearch.org/health'
Example Response
{  "status": "ok",  "wrapper": {    "uptime_seconds": 0,    "database": {      "ok": true,      "latency_ms": 0    },    "db_pool": {      "saturated": true    },    "runners": [      {        "name": "string",        "running": 0,        "seconds_since_poll": 0      }    ]  },  "backends": [    {      "model_id": "string",      "status": "live",      "gpu_shape": "string",      "breaker_state": "closed",      "breaker_consecutive_failures": 0,      "breaker_open_seconds": 0,      "in_flight": 0    }  ],  "degraded_backends": [    "string"  ],  "degraded_reasons": [    "string"  ]}

Add Sibling POST

Write a Node beside this one, under the same parent, and answer the Loom. For a root, a Loom of 2,000 Nodes, or while a Node under the parent is `queued` or `running`, 409.

Create Chat Completion POST

Not served: Base models have no chat format. Always answers 400, naming POST /v1/completions.