> 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.

# Inventory and Device Info


Register nodes, list racks and nodes, and read tray and chassis location data.

| RPC | Execution | Summary |
| --- | --- | --- |
| [`ListNodeInventory`](#listnodeinventory) | Synchronous | ListNodeInventory retrieves the complete inventory of all nodes across all racks. |
| [`CreateNodes`](#createnodes) | Synchronous | CreateNodes registers one or more new nodes in the inventory. |
| [`UpdateNode`](#updatenode) | Synchronous | UpdateNode modifies the configuration of an existing node. |
| [`DeleteNode`](#deletenode) | Synchronous | DeleteNode removes a node from the inventory. |
| [`ListRacks`](#listracks) | Synchronous | ListRacks returns a list of all rack IDs in the system. |
| [`GetNodeDeviceInfo`](#getnodedeviceinfo) | Synchronous | GetNodeDeviceInfo retrieves normalized device metadata for a specific node. |
| [`ListNodeDeviceInfoByNodeType`](#listnodedeviceinfobynodetype) | Synchronous | ListNodeDeviceInfoByNodeType retrieves normalized device metadata for all nodes of a specific type in a rack. |
| [`BatchGetNodeDeviceInfo`](#batchgetnodedeviceinfo) | Synchronous | BatchGetNodeDeviceInfo retrieves normalized device metadata for a caller-supplied list of nodes. |

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

## ListNodeInventory

`rack_manager.RackManager/ListNodeInventory`

ListNodeInventory retrieves the complete inventory of all nodes across all racks.

| 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/inventory) |
| Protobuf reference | [`ListNodeInventory`](/rms/api-reference/protobuf-reference/rack-manager/rack-manager/rack-manager-list-node-inventory) |

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

This message has no fields.

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

| Field | Type | Description |
| --- | --- | --- |
| `nodes` | repeated [`NodeInventoryInfo`](/rms/api-reference/types/rack-manager#nodeinventoryinfo) | Complete list of nodes across all racks |

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

<Accordion title="Example request">

Define `RMS_GRPCURL_FLAGS` and `RMS_ADDR` as described in [Getting started](/rms/api-reference/getting-started#call-an-rpc-with-grpcurl).

```bash
grpcurl $RMS_GRPCURL_FLAGS \
  -d '{}' \
  "$RMS_ADDR" rack_manager.RackManager/ListNodeInventory
```

The `rack_manager_client` test client offers an equivalent command:

```bash
rack_manager_client <host> <port> list_node_inventory
```

</Accordion>

## CreateNodes

`rack_manager.RackManager/CreateNodes`

CreateNodes registers one or more new nodes in the inventory. Supports batch addition of multiple nodes in a single request.

| 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/inventory) |
| Protobuf reference | [`CreateNodes`](/rms/api-reference/protobuf-reference/rack-manager/rack-manager/rack-manager-create-nodes) |

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

| Field | Type | Description |
| --- | --- | --- |
| `nodes` | [`NodeSet`](/rms/api-reference/types/rack-manager#nodeset) | List of nodes to add (batch operation supported) |

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

| Field | Type | Description |
| --- | --- | --- |
| `response` | [`OperationResponse`](/rms/api-reference/types/rack-manager#operationresponse) | - |
| `stats` | [`NodeOperationStats`](/rms/api-reference/types/rack-manager#nodeoperationstats) | Aggregate add-node counts |

**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/CreateNodes
```

The `rack_manager_client` test client offers an equivalent command:

```bash
rack_manager_client <host> <port> add_nodes <devices.csv>
```

</Accordion>

## UpdateNode

`rack_manager.RackManager/UpdateNode`

UpdateNode modifies the configuration of an existing node. Only specified fields in the update_info will be changed.

<Warning>
Not implemented. The server always returns `RETURN_CODE_FAILURE` with the message "UpdateNode not yet implemented".
</Warning>

| 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/inventory) |
| Protobuf reference | [`UpdateNode`](/rms/api-reference/protobuf-reference/rack-manager/rack-manager/rack-manager-update-node) |

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

| Field | Type | Description |
| --- | --- | --- |
| `node_id` | `string` | Target node identifier |
| `update_info` | [`NodeUpdateInfo`](/rms/api-reference/types/rack-manager#nodeupdateinfo) | Fields to update (only set fields are changed) |
| `rack_id` | `string` | Rack containing the target node |

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

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

**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 $RMS_GRPCURL_FLAGS \
  -d '{
  "node_id": "<node_id>",
  "update_info": {
    "bmc_endpoint": {
      "interface": {
        "ip_address": "<ip_address>",
        "mac_address": "<mac_address>",
        "host_name": "<host_name>"
      },
      "port": 0,
      "credentials": {
        "user_pass": {
          "username": "<username>",
          "password": "<password>"
        }
      }
    },
    "host_endpoint": {
      "interface": {
        "ip_address": "<ip_address>",
        "mac_address": "<mac_address>",
        "host_name": "<host_name>"
      },
      "port": 0,
      "credentials": {
        "user_pass": {
          "username": "<username>",
          "password": "<password>"
        }
      }
    }
  },
  "rack_id": "<rack_id>"
}' \
  "$RMS_ADDR" rack_manager.RackManager/UpdateNode
```

</Accordion>

## DeleteNode

`rack_manager.RackManager/DeleteNode`

DeleteNode removes a node from the inventory. The node must exist in the specified rack.

| 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/inventory) |
| Protobuf reference | [`DeleteNode`](/rms/api-reference/protobuf-reference/rack-manager/rack-manager/rack-manager-delete-node) |

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

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

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

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

**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/DeleteNode
```

</Accordion>

## ListRacks

`rack_manager.RackManager/ListRacks`

ListRacks returns a list of all rack IDs in the system.

| 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/inventory) |
| Protobuf reference | [`ListRacks`](/rms/api-reference/protobuf-reference/rack-manager/rack-manager/rack-manager-list-racks) |

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

This message has no fields.

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

| Field | Type | Description |
| --- | --- | --- |
| `rack_ids` | repeated `string` | List of all rack identifiers |

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

<Accordion title="Example request">

Define `RMS_GRPCURL_FLAGS` and `RMS_ADDR` as described in [Getting started](/rms/api-reference/getting-started#call-an-rpc-with-grpcurl).

```bash
grpcurl $RMS_GRPCURL_FLAGS \
  -d '{}' \
  "$RMS_ADDR" rack_manager.RackManager/ListRacks
```

The `rack_manager_client` test client offers an equivalent command:

```bash
rack_manager_client <host> <port> list_racks
```

</Accordion>

## GetNodeDeviceInfo

`rack_manager.RackManager/GetNodeDeviceInfo`

GetNodeDeviceInfo retrieves normalized device metadata for a specific node.

| 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/inventory) |
| Protobuf reference | [`GetNodeDeviceInfo`](/rms/api-reference/protobuf-reference/rack-manager/rack-manager/rack-manager-get-node-device-info) |

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

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

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

| Field | Type | Description |
| --- | --- | --- |
| `status` | [`ReturnCode`](/rms/api-reference/types/rack-manager#returncode) | RETURN_CODE_SUCCESS if retrieved, RETURN_CODE_FAILURE otherwise |
| `message` | `string` | Human-readable status or error message |
| `device_info` | [`NodeDeviceInfo`](/rms/api-reference/types/rack-manager#nodedeviceinfo) | Normalized device metadata for the 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 $RMS_GRPCURL_FLAGS \
  -d '{
  "rack_id": "<rack_id>",
  "node_id": "<node_id>"
}' \
  "$RMS_ADDR" rack_manager.RackManager/GetNodeDeviceInfo
```

The `rack_manager_client` test client offers an equivalent command:

```bash
rack_manager_client <host> <port> switch device_info <rack_id> <node_id>
```

</Accordion>

## ListNodeDeviceInfoByNodeType

`rack_manager.RackManager/ListNodeDeviceInfoByNodeType`

ListNodeDeviceInfoByNodeType retrieves normalized device metadata for all nodes of a specific type in a rack.

| 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/inventory) |
| Protobuf reference | [`ListNodeDeviceInfoByNodeType`](/rms/api-reference/protobuf-reference/rack-manager/rack-manager/rack-manager-list-node-device-info-by-node-type) |

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

| Field | Type | Description |
| --- | --- | --- |
| `rack_id` | `string` | Target rack identifier |
| `node_type` | [`NodeType`](/rms/api-reference/types/rack-manager#nodetype) | Type of nodes to query; non-UNSPECIFIED values override node_descriptor |
| `node_descriptor` | optional [`NodeDescriptor`](/rms/api-reference/types/rack-manager#nodedescriptor) | Flexible node identity used when node_type is NODE_TYPE_UNSPECIFIED |

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

| Field | Type | Description |
| --- | --- | --- |
| `status` | [`ReturnCode`](/rms/api-reference/types/rack-manager#returncode) | RETURN_CODE_SUCCESS if retrieved, RETURN_CODE_FAILURE otherwise |
| `message` | `string` | Human-readable status or error message |
| `node_device_details` | repeated [`NodeDeviceInfo`](/rms/api-reference/types/rack-manager#nodedeviceinfo) | Device metadata details grouped by node |
| `stats` | [`NodeOperationStats`](/rms/api-reference/types/rack-manager#nodeoperationstats) | Aggregate query counts |

**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 $RMS_GRPCURL_FLAGS \
  -d '{
  "rack_id": "<rack_id>",
  "node_type": "COMPUTE_GB200_NVIDIA",
  "node_descriptor": {
    "attributes": {
      "<key>": "<attributes>"
    }
  }
}' \
  "$RMS_ADDR" rack_manager.RackManager/ListNodeDeviceInfoByNodeType
```

The `rack_manager_client` test client offers an equivalent command:

```bash
rack_manager_client <host> <port> switch device_info_by_type <rack_id> <node_type>
```

</Accordion>

## BatchGetNodeDeviceInfo

`rack_manager.RackManager/BatchGetNodeDeviceInfo`

BatchGetNodeDeviceInfo retrieves normalized device metadata for a caller-supplied list of nodes. Each node is identified by its NodeInfo.

| 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/inventory) |
| Protobuf reference | [`BatchGetNodeDeviceInfo`](/rms/api-reference/protobuf-reference/rack-manager/rack-manager/rack-manager-batch-get-node-device-info) |

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

| Field | Type | Description |
| --- | --- | --- |
| `nodes` | [`NodeSet`](/rms/api-reference/types/rack-manager#nodeset) | Nodes to query |

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

| Field | Type | Description |
| --- | --- | --- |
| `status` | [`ReturnCode`](/rms/api-reference/types/rack-manager#returncode) | RETURN_CODE_SUCCESS if all requested nodes succeeded, RETURN_CODE_FAILURE otherwise |
| `message` | `string` | Batch summary; includes per-node failure details when any node fails |
| `node_device_details` | repeated [`NodeDeviceInfo`](/rms/api-reference/types/rack-manager#nodedeviceinfo) | Successful device metadata details |
| `stats` | [`NodeOperationStats`](/rms/api-reference/types/rack-manager#nodeoperationstats) | Aggregate query counts |

**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/BatchGetNodeDeviceInfo
```

The `rack_manager_client` test client offers an equivalent command:

```bash
rack_manager_client <host> <port> switch device_info_batch <nodes.csv>
```

</Accordion>