> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs.nvidia.com/rms/llms.txt.

# Power Control


Read and change node power state, and run the coordinated rack cold reboot.

| RPC | Execution | Summary |
| --- | --- | --- |
| [`SetPowerState`](#setpowerstate) | Synchronous | SetPowerState controls the power state of a single node. |
| [`BatchSetPowerState`](#batchsetpowerstate) | Synchronous | BatchSetPowerState controls power for caller-supplied nodes without requiring them to be registered in RMS inventory. |
| [`GetPowerState`](#getpowerstate) | Synchronous | GetPowerState retrieves the current power state of a node. |
| [`BatchGetPowerState`](#batchgetpowerstate) | Synchronous | BatchGetPowerState retrieves the current power state for caller-supplied nodes without requiring them to be registered in RMS inventory. |
| [`ExecuteColdReboot`](#executecoldreboot) | Asynchronous, one job | Run the full rack cold reboot sequence on caller-supplied nodes as one asynchronous job. |

For definitions of execution modes and job polling, refer to [Async jobs](/rms/api-reference/async-jobs).

## SetPowerState

`rack_manager.RackManager/SetPowerState`

SetPowerState controls the power state of a single node. Supports power on, power off, and power reset operations.

| Property | Value |
| --- | --- |
| Execution | **Synchronous.** The response carries the full result. |
| Target | **Registered inventory.** Reads or changes nodes and racks stored in RMS inventory. |
| Behavior details | [Operations guide](/rms/documentation/operations/power-control) |
| Protobuf reference | [`SetPowerState`](/rms/api-reference/protobuf-reference/rack-manager/rack-manager/rack-manager-set-power-state) |

**Request:** [`SetPowerStateRequest`](/rms/api-reference/types/rack-manager#setpowerstaterequest)

| Field | Type | Description |
| --- | --- | --- |
| `node_id` | `string` | Target node identifier |
| `operation` | [`PowerOperation`](/rms/api-reference/types/rack-manager#poweroperation) | Power operation to perform |
| `rack_id` | `string` | Rack containing the target node |

**Response:** [`SetPowerStateResponse`](/rms/api-reference/types/rack-manager#setpowerstateresponse)

| Field | Type | Description |
| --- | --- | --- |
| `status` | [`ReturnCode`](/rms/api-reference/types/rack-manager#returncode) | RETURN_CODE_SUCCESS if power operation completed, RETURN_CODE_FAILURE otherwise |

**Errors:** Refer to [Conventions](/rms/api-reference/conventions#error-handling) for how RMS reports failures. This RPC also returns these gRPC errors:

- `INVALID_ARGUMENT`: The `operation` value is not a valid `PowerOperation`.

<Accordion title="Example request">

Replace each `<placeholder>` with a real value. Define `RMS_GRPCURL_FLAGS` and `RMS_ADDR` as described in [Getting started](/rms/api-reference/getting-started#call-an-rpc-with-grpcurl).

```bash
grpcurl $RMS_GRPCURL_FLAGS \
  -d '{
  "node_id": "<node_id>",
  "operation": "POWER_OPERATION_OFF",
  "rack_id": "<rack_id>"
}' \
  "$RMS_ADDR" rack_manager.RackManager/SetPowerState
```

The `rack_manager_client` test client offers an equivalent command:

```bash
rack_manager_client <host> <port> power <rack_id> <node_id> <operation>
```

</Accordion>

## BatchSetPowerState

`rack_manager.RackManager/BatchSetPowerState`

BatchSetPowerState controls power for caller-supplied nodes without requiring them to be registered in RMS inventory.

| Property | Value |
| --- | --- |
| Execution | **Synchronous.** The response carries the full result. |
| Target | **Caller-supplied nodes.** The request carries each node's endpoints and credentials, and RMS does not store them. |
| Behavior details | [Operations guide](/rms/documentation/operations/power-control) |
| Protobuf reference | [`BatchSetPowerState`](/rms/api-reference/protobuf-reference/rack-manager/rack-manager/rack-manager-batch-set-power-state) |

**Request:** [`BatchSetPowerStateRequest`](/rms/api-reference/types/rack-manager#batchsetpowerstaterequest)

| Field | Type | Description |
| --- | --- | --- |
| `nodes` | [`NodeSet`](/rms/api-reference/types/rack-manager#nodeset) | Caller-supplied node connection details |
| `operation` | [`PowerOperation`](/rms/api-reference/types/rack-manager#poweroperation) | Power operation to perform |

**Response:** [`BatchSetPowerStateResponse`](/rms/api-reference/types/rack-manager#batchsetpowerstateresponse)

| Field | Type | Description |
| --- | --- | --- |
| `response` | [`NodeBatchResponse`](/rms/api-reference/types/rack-manager#nodebatchresponse) | - |

**Errors:** Refer to [Conventions](/rms/api-reference/conventions#error-handling) for how RMS reports failures. This RPC also returns these gRPC errors:

- `INVALID_ARGUMENT`: The `operation` value is not a valid `PowerOperation`.

<Accordion title="Example request">

Replace each `<placeholder>` with a real value. Define `RMS_GRPCURL_FLAGS` and `RMS_ADDR` as described in [Getting started](/rms/api-reference/getting-started#call-an-rpc-with-grpcurl).

```bash
grpcurl $RMS_GRPCURL_FLAGS \
  -d '{
  "nodes": {
    "nodes": [
      {
        "node_id": "<node_id>",
        "rack_id": "<rack_id>",
        "type": "COMPUTE_GB200_NVIDIA",
        "bmc_endpoint": {
          "interface": {
            "ip_address": "<ip_address>",
            "mac_address": "<mac_address>",
            "host_name": "<host_name>"
          },
          "port": 0,
          "credentials": {
            "user_pass": {
              "username": "<username>",
              "password": "<password>"
            }
          }
        },
        "host_endpoint": {
          "interface": {
            "ip_address": "<ip_address>",
            "mac_address": "<mac_address>",
            "host_name": "<host_name>"
          },
          "port": 0,
          "credentials": {
            "user_pass": {
              "username": "<username>",
              "password": "<password>"
            }
          }
        },
        "node_descriptor": {
          "attributes": {
            "<key>": "<attributes>"
          }
        },
        "additional_host_endpoints": [
          {
            "interface": {
              "ip_address": "<ip_address>",
              "mac_address": "<mac_address>",
              "host_name": "<host_name>"
            },
            "port": 0,
            "credentials": {
              "user_pass": {
                "username": "<username>",
                "password": "<password>"
              }
            }
          }
        ]
      }
    ]
  },
  "operation": "POWER_OPERATION_OFF"
}' \
  "$RMS_ADDR" rack_manager.RackManager/BatchSetPowerState
```

The `rack_manager_client` test client offers an equivalent command:

```bash
rack_manager_client <host> <port> batch_set_power_state <nodes.csv> <operation>
```

</Accordion>

## GetPowerState

`rack_manager.RackManager/GetPowerState`

GetPowerState retrieves the current power state of a node. Returns the current state (ON, OFF, or UNKNOWN).

| Property | Value |
| --- | --- |
| Execution | **Synchronous.** The response carries the full result. |
| Target | **Registered inventory.** Reads or changes nodes and racks stored in RMS inventory. |
| Behavior details | [Operations guide](/rms/documentation/operations/power-control) |
| Protobuf reference | [`GetPowerState`](/rms/api-reference/protobuf-reference/rack-manager/rack-manager/rack-manager-get-power-state) |

**Request:** [`GetPowerStateRequest`](/rms/api-reference/types/rack-manager#getpowerstaterequest)

| Field | Type | Description |
| --- | --- | --- |
| `node_id` | `string` | Target node identifier |
| `rack_id` | `string` | Rack containing the target node |

**Response:** [`GetPowerStateResponse`](/rms/api-reference/types/rack-manager#getpowerstateresponse)

| Field | Type | Description |
| --- | --- | --- |
| `status` | [`ReturnCode`](/rms/api-reference/types/rack-manager#returncode) | RETURN_CODE_SUCCESS if query completed, RETURN_CODE_FAILURE otherwise |
| `node_id` | `string` | Node identifier (echoed from request) |
| `pstate` | `string` | Current power state: "ON", "OFF", or "UNKNOWN" |
| `rack_id` | `string` | Rack identifier (echoed from request) |

**Errors:** Refer to [Conventions](/rms/api-reference/conventions#error-handling) for how RMS reports failures.

<Accordion title="Example request">

Replace each `<placeholder>` with a real value. Define `RMS_GRPCURL_FLAGS` and `RMS_ADDR` as described in [Getting started](/rms/api-reference/getting-started#call-an-rpc-with-grpcurl).

```bash
grpcurl $RMS_GRPCURL_FLAGS \
  -d '{
  "node_id": "<node_id>",
  "rack_id": "<rack_id>"
}' \
  "$RMS_ADDR" rack_manager.RackManager/GetPowerState
```

The `rack_manager_client` test client offers an equivalent command:

```bash
rack_manager_client <host> <port> status <rack_id> <node_id>
```

</Accordion>

## BatchGetPowerState

`rack_manager.RackManager/BatchGetPowerState`

BatchGetPowerState retrieves the current power state for caller-supplied nodes without requiring them to be registered in RMS inventory. Returns per-node power state along with batch status.

| Property | Value |
| --- | --- |
| Execution | **Synchronous.** The response carries the full result. |
| Target | **Caller-supplied nodes.** The request carries each node's endpoints and credentials, and RMS does not store them. |
| Behavior details | [Operations guide](/rms/documentation/operations/power-control) |
| Protobuf reference | [`BatchGetPowerState`](/rms/api-reference/protobuf-reference/rack-manager/rack-manager/rack-manager-batch-get-power-state) |

**Request:** [`BatchGetPowerStateRequest`](/rms/api-reference/types/rack-manager#batchgetpowerstaterequest)

| Field | Type | Description |
| --- | --- | --- |
| `nodes` | [`NodeSet`](/rms/api-reference/types/rack-manager#nodeset) | Caller-supplied node connection details |

**Response:** [`BatchGetPowerStateResponse`](/rms/api-reference/types/rack-manager#batchgetpowerstateresponse)

| Field | Type | Description |
| --- | --- | --- |
| `response` | [`NodeBatchResponse`](/rms/api-reference/types/rack-manager#nodebatchresponse) | - |
| `node_power_states` | repeated [`NodePowerState`](/rms/api-reference/types/rack-manager#nodepowerstate) | Power state for each successfully queried node |

**Errors:** Refer to [Conventions](/rms/api-reference/conventions#error-handling) for how RMS reports failures.

<Accordion title="Example request">

Replace each `<placeholder>` with a real value. Define `RMS_GRPCURL_FLAGS` and `RMS_ADDR` as described in [Getting started](/rms/api-reference/getting-started#call-an-rpc-with-grpcurl).

```bash
grpcurl $RMS_GRPCURL_FLAGS \
  -d '{
  "nodes": {
    "nodes": [
      {
        "node_id": "<node_id>",
        "rack_id": "<rack_id>",
        "type": "COMPUTE_GB200_NVIDIA",
        "bmc_endpoint": {
          "interface": {
            "ip_address": "<ip_address>",
            "mac_address": "<mac_address>",
            "host_name": "<host_name>"
          },
          "port": 0,
          "credentials": {
            "user_pass": {
              "username": "<username>",
              "password": "<password>"
            }
          }
        },
        "host_endpoint": {
          "interface": {
            "ip_address": "<ip_address>",
            "mac_address": "<mac_address>",
            "host_name": "<host_name>"
          },
          "port": 0,
          "credentials": {
            "user_pass": {
              "username": "<username>",
              "password": "<password>"
            }
          }
        },
        "node_descriptor": {
          "attributes": {
            "<key>": "<attributes>"
          }
        },
        "additional_host_endpoints": [
          {
            "interface": {
              "ip_address": "<ip_address>",
              "mac_address": "<mac_address>",
              "host_name": "<host_name>"
            },
            "port": 0,
            "credentials": {
              "user_pass": {
                "username": "<username>",
                "password": "<password>"
              }
            }
          }
        ]
      }
    ]
  }
}' \
  "$RMS_ADDR" rack_manager.RackManager/BatchGetPowerState
```

The `rack_manager_client` test client offers an equivalent command:

```bash
rack_manager_client <host> <port> batch_get_power_state <nodes.csv>
```

</Accordion>

## ExecuteColdReboot

`rack_manager.RackManager/ExecuteColdReboot`

ExecuteColdReboot orchestrates the full DGX GB200 rack COLD reboot sequence on caller-supplied nodes (no inventory registration required): gracefully shut down the compute trays over BMC, power off the power shelves, wait for capacitor discharge, power the shelves back on, wait for the compute/switch tray BMCs and switch NVOS to come up, then power on the compute nodes. Returns a coordinator job ID immediately; poll GetJobStatus on it: while running, execution_state is RUNNING and state_description reports the current stage by its rack-reboot runbook step number(s) (e.g. "Rack reboot step 7-8: ..."); on completion or failure the full per-stage stage timeline is written to result_json. Runbook steps 9 (switch fabric provisioning) and 13 (post-reboot host/GPU verification) are out of scope. The sequence is fail-fast with no automated rollback.

<Warning>
The sequence stops at the first failed stage and does not roll back. The response `job_id` identifies the coordinator job.
</Warning>

| Property | Value |
| --- | --- |
| Execution | **Asynchronous, one job.** The response returns a job ID and the work continues after the call returns. |
| Job tracking | Poll [`GetJobStatus`](/rms/api-reference/rpcs/utility-jobs#getjobstatus) with the job ID in `response.job_id`. |
| Job result | `result_json` in the job status holds the full per-stage timeline of the reboot sequence, written on completion or failure. |
| Target | **Caller-supplied nodes.** The request carries each node's endpoints and credentials, and RMS does not store them. |
| Behavior details | [Operations guide](/rms/documentation/operations/power-control) |
| Protobuf reference | [`ExecuteColdReboot`](/rms/api-reference/protobuf-reference/rack-manager/rack-manager/rack-manager-execute-cold-reboot) |

**Request:** [`ExecuteColdRebootRequest`](/rms/api-reference/types/rack-manager#executecoldrebootrequest)

| Field | Type | Description |
| --- | --- | --- |
| `nodes` | [`NodeSet`](/rms/api-reference/types/rack-manager#nodeset) | Participating devices: compute + power-shelf + switch trays (full NodeInfo) |
| `domain` | optional `string` | Optional NVL domain selecting switch mTLS material for the NVOS-readiness probe |
| `prestage_shutdown_floor_seconds` | optional `uint32` | Min wait after issuing the compute graceful shutdown before probing (default 0) |
| `prestage_shutdown_poll_max_seconds` | optional `uint32` | Max wait for compute to power down before the shelves are cut (default 300) |
| `discharge_floor_seconds` | optional `uint32` | Min wait after shelf power-off before probing (default 180) |
| `discharge_poll_max_seconds` | optional `uint32` | Max wait for BMCs to become unreachable (default 600) |
| `bmc_up_floor_seconds` | optional `uint32` | Min wait after shelf power-on before probing (default 120) |
| `bmc_up_poll_max_seconds` | optional `uint32` | Max wait for BMCs to become reachable (default 600) |
| `nvos_hello_backoff_seconds` | optional `uint32` | Base delay for the exponential backoff between switch nvue-hello attempts (default 30) |
| `compute_boot_wait_seconds` | optional `uint32` | Fixed wait for the compute OS to boot after power-on (default 300); the BMC cannot report OS readiness |
| `poll_interval_seconds` | optional `uint32` | Poll cadence for all stages (default 15) |

**Response:** [`ExecuteColdRebootResponse`](/rms/api-reference/types/rack-manager#executecoldrebootresponse)

| Field | Type | Description |
| --- | --- | --- |
| `response` | [`NodeBatchResponse`](/rms/api-reference/types/rack-manager#nodebatchresponse) | response.job_id is the coordinator job ID for GetJobStatus |

**Errors:** Refer to [Conventions](/rms/api-reference/conventions#error-handling) for how RMS reports failures.

<Accordion title="Example request">

Replace each `<placeholder>` with a real value. Define `RMS_GRPCURL_FLAGS` and `RMS_ADDR` as described in [Getting started](/rms/api-reference/getting-started#call-an-rpc-with-grpcurl).

```bash
grpcurl $RMS_GRPCURL_FLAGS \
  -d '{
  "nodes": {
    "nodes": [
      {
        "node_id": "<node_id>",
        "rack_id": "<rack_id>",
        "type": "COMPUTE_GB200_NVIDIA",
        "bmc_endpoint": {
          "interface": {
            "ip_address": "<ip_address>",
            "mac_address": "<mac_address>",
            "host_name": "<host_name>"
          },
          "port": 0,
          "credentials": {
            "user_pass": {
              "username": "<username>",
              "password": "<password>"
            }
          }
        },
        "host_endpoint": {
          "interface": {
            "ip_address": "<ip_address>",
            "mac_address": "<mac_address>",
            "host_name": "<host_name>"
          },
          "port": 0,
          "credentials": {
            "user_pass": {
              "username": "<username>",
              "password": "<password>"
            }
          }
        },
        "node_descriptor": {
          "attributes": {
            "<key>": "<attributes>"
          }
        },
        "additional_host_endpoints": [
          {
            "interface": {
              "ip_address": "<ip_address>",
              "mac_address": "<mac_address>",
              "host_name": "<host_name>"
            },
            "port": 0,
            "credentials": {
              "user_pass": {
                "username": "<username>",
                "password": "<password>"
              }
            }
          }
        ]
      }
    ]
  },
  "domain": "<domain>",
  "prestage_shutdown_floor_seconds": 0,
  "prestage_shutdown_poll_max_seconds": 0,
  "discharge_floor_seconds": 0,
  "discharge_poll_max_seconds": 0,
  "bmc_up_floor_seconds": 0,
  "bmc_up_poll_max_seconds": 0,
  "nvos_hello_backoff_seconds": 0,
  "compute_boot_wait_seconds": 0,
  "poll_interval_seconds": 0
}' \
  "$RMS_ADDR" rack_manager.RackManager/ExecuteColdReboot
```

</Accordion>