Support API Key Authentication

Reset pod GPUs

Changed on

Start a reset of a running pod's GPUs (DAH-4658): the pod's container is stopped, the GPUs are reset and the container is started again. Every process in the pod stops; files are kept. gpu_indices are the pod's own indices (as nvidia-smi shows them inside the pod); omit them to reset every GPU of the pod. The answer is status: running with a request_id: poll GET /pods/{id}/reset-gpu/{request_id} for the result. Errors name their reason in error.code.

post/pods/{id}/reset-gpu

Request

  • Base URL
    https://lium.io/api
  • URL
    https://lium.io/api/pods/{id}/reset-gpu
  • Auth
    • HTTP bearer
    • or API key in header X-API-Key

Path parameters

idstring uuid required

Request body

gpu_indicesinteger[] nullable

Response

The reset started; poll its request_id

pod_idstring uuid required
request_idstring required
status'running' | 'done' | 'failed' | 'timeout' required
okboolean nullable
container_restartedboolean nullable

Changes