R
RunPod•3w ago
Bersaelor

How to set ContainerRegistryAuth for `podRentInterruptable`

When renting interruptable pods using GraphQL, like so: 
mutation { 
  podRentInterruptable( input: { 
    bidPerGpu: 0.09, 
    cloudType: COMMUNITY, 
    gpuCount: 1, 
    volumeInGb: 4, 
    containerDiskInGb: 4, 
    minVcpuCount: 2, 
    minMemoryInGb: 8, 
    gpuTypeId: "NVIDIA GeForce RTX 3080 Ti", 
    name: "gpu-dev-ovh-dev-0-0", 
    imageName: "efnats/chia-recompute:develop", 
    containerRegistryAuthId: "<id>",
    dockerArgs: "", 
    ports: "", 
    volumeMountPath: "/workspace", 
    env: [
      { key: "PROXY_IP", value: "51.195.44.53"},
      { key: "INSTANCE_NAME", value: "gpu-dev-ovh-dev-0-0"}
    ] 
  } ) { 
      id
      imageName 
      env 
      machineId 
      machine { 
        podHostId 
      } 
    } 
}
mutation { 
  podRentInterruptable( input: { 
    bidPerGpu: 0.09, 
    cloudType: COMMUNITY, 
    gpuCount: 1, 
    volumeInGb: 4, 
    containerDiskInGb: 4, 
    minVcpuCount: 2, 
    minMemoryInGb: 8, 
    gpuTypeId: "NVIDIA GeForce RTX 3080 Ti", 
    name: "gpu-dev-ovh-dev-0-0", 
    imageName: "efnats/chia-recompute:develop", 
    containerRegistryAuthId: "<id>",
    dockerArgs: "", 
    ports: "", 
    volumeMountPath: "/workspace", 
    env: [
      { key: "PROXY_IP", value: "51.195.44.53"},
      { key: "INSTANCE_NAME", value: "gpu-dev-ovh-dev-0-0"}
    ] 
  } ) { 
      id
      imageName 
      env 
      machineId 
      machine { 
        podHostId 
      } 
    } 
}
containerRegistryAuthId doesn't seem to be part of the PodRentInterruptableInput https://graphql-spec.runpod.io/#definition-PodRentInterruptableInput specification. How do we rent an interruptible pod where the image is private on DockerHub?
2 Replies
nerdylive
nerdylive•3w ago
How about using template ID?
Bersaelor
BersaelorOP•3w ago
you mean, creating the template via the web console and then using the templateID in the graphql call? The env variables I send to the instance are dynamic, so I guess I would have to create a new template every time? ah, or I can just modify the template before I create a new pod, that would work too, using graphql

Did you find this page helpful?