Skip to main content
POST
Make one recovery attempt now on every cluster node that is not running, regardless of the policy's cooldown and cap (no new hardware, no new rate)

Authorizations

Authorization
string
header
required

A management token: minted by the OAuth flow (aireserve login, or an MCP client's Authenticate step) or on the portal Keys page under Management Keys. Inference API keys are refused.

Body

application/json
clusterId
string<uuid>
required

Cluster id from list_gpu_clusters

Pattern: ^([0-9a-fA-F]{8}-[0-9a-fA-F]{4}-[1-8][0-9a-fA-F]{3}-[89abAB][0-9a-fA-F]{3}-[0-9a-fA-F]{12}|00000000-0000-0000-0000-000000000000|ffffffff-ffff-ffff-ffff-ffffffffffff)$
clientId
string<uuid>

Target organization id — platform admins only; omit otherwise

Pattern: ^([0-9a-fA-F]{8}-[0-9a-fA-F]{4}-[1-8][0-9a-fA-F]{3}-[89abAB][0-9a-fA-F]{3}-[0-9a-fA-F]{12}|00000000-0000-0000-0000-000000000000|ffffffff-ffff-ffff-ffff-ffffffffffff)$

Response

Executed successfully (ok: true).

status
enum<string>
required
Available options:
executed
ok
boolean
required

Whether the operation succeeded.

summary
string
required

The consent line describing what was done, built from resolved labels.

message
string
required

Human-readable result line.

data
any

Operation payload (reads) or detail worth relaying (writes).

secret
string

A one-time secret the operation minted (API key, credential). Shown once; store it now.

sensitive
enum<boolean>

Present exactly when secret is: redact this response from logs.

Available options:
true