NVIDIA UFM Enterprise REST API Guide v6.23.1

NVLink REST API

The REST API is grouped into the following categories:

  • App REST APIs: General plugin information

  • Managed Domains APIs: Managing domain connections

  • Inventory APIs: Querying hardware and configuration details

Get Version

  • Description: Get a version of NVLink Aggregator

  • URL: GET /ufmRestV2/plugin/nmxaggr/v1/app/version

  • Responses:

    • Status: 200 - OK

      • Description: Version is successfully returned

      • Content:

        Copy
        Copied!
                    

        { "version": "0.2.9" }

Get Managed Domains

  • Description: Get a list of managed domains

  • URL: GET /ufmRestV2/plugin/nmxaggr/v1/managed_domains

  • Responses:

    • 200 - OK

      • Description: List of managed domains is successfully returned

      • Content:

        Copy
        Copied!
                    

        { "managed_domains" : [ { "domain_uuid" : "4b47fa0f-7dbc-474b-b521-fc3ffc6dc7df", "host" : "nvlink-01", "controller_port" : 9370, "telemetry_port" : 9351     }, ... ] }

Add Managed Domains

  • Description: Add one or more new managed domains

  • URL: POST /ufmRestV2/plugin/nmxaggr/v1/managed_domains/add

  • Request Body:

    Parameter

    Type

    Description

    Required

    Possible values

    Default Value

    host

    string

    Hostname or IP of the switch running NMX Controller. Both hostname and IP can contain range expressions

    true

    controller_port

    integer

    Port number of NMX Controller gRPC API server

    false

    9370

    telemetry_port

    integer

    Port number of NMX Telemetry gRPC API server

    false

    9351

    Copy
    Copied!
                

    [ {"host" : "10.245.16.133", "controller_port" : 6901, "telemetry_port" : 9351}, {"host" : "10.245.[16,17].[134-137]"}, {"host" : "nvlink-[001-32]", "telemetry_port" : 9352} ]

  • Responses:

    • Status: 200 - OK

      • Description: Operation is performed. The response body contains a list of operation statuses for each domain in a request

      • Content:

        Parameter

        Type

        Description

        host

        string

        A hostname or an IP of a domain

        success

        boolean

        Whether an operation has succeeded for this domain or not

        reason

        string

        A reason for an operation failure

        Copy
        Copied!
                    

        [ {"host" : "nvlink-001", "success" : true, "reason" : ""},     {"host" : "nvlink-002", "success" : false, "reason" : "already managed"},     {"host" : "nvlink-003", "success" : false, "reason" : "unreachable"} ]

Remove Managed Domains

  • Description: Remove one or more existing managed domains

  • URL:POST /ufmRestV2/plugin/nmxaggr/v1/managed_domains/remove

  • Request Body:

    Parameter

    Type

    Description

    Required

    Possible values

    Default Value

    host

    string

    Hostname or IP of the switch running NMX Controller. Both hostname and IP can contain range expressions

    true

    Copy
    Copied!
                

    [ {"host" : "10.245.16.133"}, {"host" : "10.245.[16,17].[134-137]"}, {"host" : "nvlink-[001-32]"} ]

  • Responses:

    • Status: 200 - OK

      • Description: Operation is performed. The response body contains a list of operation statuses for each domain in a request

      • Content:

        Parameter

        Type

        Description

        host

        string

        A hostname or an IP of a domain

        success

        boolean

        Whether an operation has succeeded for this domain or not

        reason

        string

        A reason for an operation failure

        Copy
        Copied!
                    

        [ {"host" : "nvlink-001", "success" : true, "reason" : ""},     {"host" : "nvlink-002", "success" : false, "reason" : "isn't managed"} ]

Get Domains

  • Description: Gets a list of domains

  • URL: GET /ufmRestV2/plugin/nmxaggr/v1/inventory/domains

  • Parameters:

    Parameter

    Type

    Description

    Required

    Example

    Possible Values

    domain_uuid

    string

    Return only one domain with this UUID

    false

    ?domain_uuid=2e469d41-3102-485c-aa8b-58ba3c97f6bb

    domain_uuids

    list of strings

    Return only domains with UUIDs provided in a comma separated list

    false

    ?domain_uuids=2e469d41-3102-485c-aa8b-58ba3c97f6bb,d7c35619-fa03-4bb7-bce2-e4e37614b38b

    health

    string

    Filter returned domains by a value of the "health" property

    false

    ?health=unhealthy

    healthy, degraded , unhealthy, unreachable

  • Responses:

    • Status: 200 - OK

      • Description: List of domains is successfully returned

      • Content:

        Copy
        Copied!
                    

        { "domains" : [ { "name" : "locujoko", "domain_uuid" : "4b47fa0f-7dbc-474b-b521-fc3ffc6dc7df", "controller_address" : {"host" : "10.245.16.133", "port" : 6901}, "telemetry_address" : {"host" : "10.245.16.133", "port" : 9351},       "counts" : { "num_gpus" : 36, "num_switches" : 18, "num_ports" : 1296, "num_links" : 648 }, "properties" : { "max_compute_nodes" : 9, "max_compute_nodes_per_chassis" : 9, "max_gpus_per_compute_node" : 4, "max_gpu_nvlinks" : 18, "max_switch_nodes" : 9, "max_switch_nodes_per_chassis" : 9, "max_switches_per_switch_node" : 2, "max_switch_nvlinks" : 36, "max_ports" : 1296, "max_partitions" : 36, "min_gpus_per_partition" : 0, "max_alids" : 2, "max_multicast_groups" : 1024, "line_rate" : 25000 }, "controller" : { "app_uuid" : "0eb9b4a7-62fc-4ce7-b51a-50b6fa9f1d50", "app_ver" : "1.1.0_2025-05-09_21-04", "host_os_details" : "Ubuntu 22.04.5 LTS", "components_ver" : { "fib" : "1.1.0_2025-05-09_21-04", "gfm" : "R570.148", "sm" : "2025.03.6" }, "capabilities" : [ "fib", "gfm", "gw-api", "sm" ] },      "telemetry_controller": { "app_uuid": "0ae2c6c7-6d90-410f-8e54-d9985eee95a0", "app_ver": "3.0.4", "host_os_details": "Ubuntu 22.04.5 LTS", "components_ver": { "clusterminder": "1.1.10", "envoy": "1.33.0", "gnmi-aggregator": "1.15.0", "gnmi_plugin": "gnmi_telemetry_1.3.4-0.tar", "nmx-connector": "1.15.0", "nvl-telemetry": "1.21.4" }, "capabilities": [ "nvl telemetry", "gnmi aggregation", "syslog aggregation", "redfish_aggregation" ] }, "telemetry_config": { "enabled": true, "update_period": 60, "collect": { "device": { "gpu": true, "switch": true }, "down_ports": false }, "otlp": { "counter_set": "", "field_set": "", "target": "", "inbound_queue": 10000 }, "remote_write": { "counter_set": "", "field_set": "", "target": "", "inbound_queue": 10000 } },       "health" : "Healthy", "state" : "Configured",      "config_status" : "CONFIG_DONE",       "controller_health" : "Healthy",      "telemetry_health": "Healthy", "last_updated" : "2025-05-21T13:12:11+03:00" }, ... ] }

    • Status: 404 - Not Found

      • Description: The domain with one of specified UUIDs could not be found

      • Content:

        Copy
        Copied!
                    

        { "error": "the domain with the uuid '4b47fa0f-7dbc-474b-b521-fc3ffc6dc7da' could not be found" }

Get Compute Nodes

  • Description: Get a list of compute nodes

  • URL: GET /ufmRestV2/plugin/nmxaggr/v1/inventory/compute_nodes

  • Parameters:

    Parameter

    Type

    Description

    Required

    Example

    Possible Values

    domain_uuid

    string

    Return only compute nodes belonging to the domain with this UUID

    false

    ?domain_uuid=2e469d41-3102-485c-aa8b-58ba3c97f6bb

    domain_uuids

    list of strings

    Return only compute nodes belonging to domains with UUIDs provided in a comma separated list

    false

    ?domain_uuids=2e469d41-3102-485c-aa8b-58ba3c97f6bb,d7c35619-fa03-4bb7-bce2-e4e37614b38b

    health

    string

    Filter returned compute nodes by a value of the "health" property

    false

    ?health=unhealthy

    healthy, degraded, unhealthy

    is_allocated

    boolean

    If true, allocated compute nodes will be returned, otherwise free compute nodes will be returned

    false

    ?is_allocated=true

  • Responses:

    • Status: 200 - OK

      • Description: List of compute nodes is successfully returned

      • Content:

        Copy
        Copied!
                    

        { "compute_nodes": [ {       "domain_name": "haxazeko", "domain_uuid": "3bc0511c-6ac2-428b-b546-843d3cad0f3f", "name": "haxazeko.chas-01.slot-03", "location": { "chassis_id": 1, "slot_id": 3 }, "location_info": { "chassis_serial_number": "27XYZ27000001", "tray_index": 2 }, "partition_ids": [ "0x7ffe" ], "last_updated": "2025-10-21T21:09:28+03:00", "num_gpus": 4, "health": "Healthy"     }, . . . ] }

    • Status: 404 - Not Found

      • Description: The domain with one of specified UUIDs could not be found

      • Content:

        Copy
        Copied!
                    

        { "error": "the domain with the uuid '4b47fa0f-7dbc-474b-b521-fc3ffc6dc7da' could not be found" }

Get Switch Nodes

  • Description: Get a list of switch nodes

  • URL: GET /ufmRestV2/plugin/nmxaggr/v1/inventory/switch_nodes

  • Parameters:

    Parameter

    Type

    Description

    Required

    Example

    Possible Values

    domain_uuid

    string

    Return only switch nodes belonging to the domain with this UUID

    false

    ?domain_uuid=2e469d41-3102-485c-aa8b-58ba3c97f6bb

    domain_uuids

    list of strings

    Return only switch nodes belonging to domains with UUIDs provided in a comma separated list

    false

    ?domain_uuids=2e469d41-3102-485c-aa8b-58ba3c97f6bb,d7c35619-fa03-4bb7-bce2-e4e37614b38b

    health

    string

    Filter returned switch nodes by a value of the "health" property

    false

    ?health=unhealthy

    healthy, degraded, unhealthy

  • Responses:

    • Status: 200 - OK

      • Description: List of switch nodes is successfully returned

      • Content:

        Copy
        Copied!
                    

        { "switch_nodes": [ {       "domain_name": "cymovewe", "domain_uuid": "d1d652ea-119e-4ffb-9f22-7023deba5c42", "name": "cymovewe.chas-01.slot-08", "location": { "chassis_id": 1, "slot_id": 8 }, "location_info": { "chassis_serial_number": "27XYZ27000001", "tray_index": 0 }, "partition_ids": [], "last_updated": "2025-10-21T21:13:59+03:00", "num_switches": 2, "health": "Healthy"     }, . . . ] }

    • Status: 404 - Not Found

      • Description: The domain with one of specified UUIDs could not be found

      • Content:

        Copy
        Copied!
                    

        { "error": "the domain with the uuid '4b47fa0f-7dbc-474b-b521-fc3ffc6dc7da' could not be found" }

Get GPUs

  • Description: Gets a list of GPUs

  • URL: GET /ufmRestV2/plugin/nmxaggr/v1/inventory/gpus

  • Parameters:

    Parameter

    Type

    Description

    Required

    Example

    Possible Values

    domain_uuid

    string

    Return only GPUs belonging to the domain with this UUID

    false

    ?domain_uuid=2e469d41-3102-485c-aa8b-58ba3c97f6bb

    domain_uuids

    list of strings

    Return only GPUs belonging to domains with UUIDs provided in a comma separated list

    false

    ?domain_uuids=2e469d41-3102-485c-aa8b-58ba3c97f6bb,d7c35619-fa03-4bb7-bce2-e4e37614b38b

    health

    string

    Filter returned GPUs by a value of the "health" property

    false

    ?health=unhealthy

    healthy, degraded, unhealthy

  • Responses:

    • Status: 200 - OK

      • Description: List of GPUs is successfully returned

      • Content:

        Copy
        Copied!
                    

        { "gpus": [ { "domain_name": "locujoko", "domain_uuid": "4b47fa0f-7dbc-474b-b521-fc3ffc6dc7df", "domain_controller_address": { "host": "10.245.16.133", "port": 6901 }, "last_updated": "2025-05-21T16:42:14+03:00", "name": "locujoko.chas-01.cmp-node-07.gpu-01", "location": { "chassis_id": 1, "slot_id": 16, "host_id": 1 }, "location_info": { "chassis_serial_number": "27XYZ27000001", "tray_index": 6 }, "description": "GB100 Nvidia Technologies 24", "device_uid": "0a48-fc18-c7cb-a43c", "system_uid": "0c64-56a8-0d9e-37c7", "topology_id": 129, "price_id": 10496, "num_ports": 18, "vendor_id": 713, "device_id": 1, "partition_ids": [ 32766 ], "health": "Healthy", "alids": [ 1048 ] }, . . . ] }

    • Status: 404 - Not Found

      • Description: The domain with one of specified UUIDs could not be found

      • Content:

        Copy
        Copied!
                    

        { "error": "the domain with the uuid '4b47fa0f-7dbc-474b-b521-fc3ffc6dc7da' could not be found" }

Get Switches

  • Description: Gets a list of switches

  • URL: GET /ufmRestV2/plugin/nmxaggr/v1/inventory/switches

  • Parameters:

    Parameter

    Type

    Description

    Required

    Example

    Possible Values

    domain_uuid

    string

    Return only switches belonging to the domain with this UUID

    false

    ?domain_uuid=2e469d41-3102-485c-aa8b-58ba3c97f6bb

    domain_uuids

    list of strings

    Return only switches belonging to domains with UUIDs provided in a comma separated list

    false

    ?domain_uuids=2e469d41-3102-485c-aa8b-58ba3c97f6bb,d7c35619-fa03-4bb7-bce2-e4e37614b38b

    health

    string

    Filter returned switches by a value of the "health" property

    false

    ?health=unhealthy

    healthy, degraded, unhealthy

  • Responses:

    • Status: 200 - OK

      • Description: List of switches is successfully returned

      • Content:

        Copy
        Copied!
                    

        { "switches" : [ { "domain_name" : "locujoko", "domain_uuid" : "4b47fa0f-7dbc-474b-b521-fc3ffc6dc7df", "domain_controller_address" : {"host" : "10.245.16.133", "port" : 6901}, "last_updated" : "2025-05-21T16:42:14+03:00", "name" : "locujoko.chas-01.swt-node-01.swt-01", "location" : {"chassis_id" : 1, "slot_id" : 13, "host_id" : 1}, "location_info" : {"chassis_serial_number" : "27XYZ27000001", "tray_index" : 0}, "description" : "MF0;mc-gb-nvl-020-001-switch:N5110_LD/U 16", "device_uid" : "0000-0002-5133-445a", "system_uid" : "0d6d-e7fc-d5da-13b0", "topology_id" : 129, "price_id" : 54004, "num_ports" : 72, "vendor_id" : 713, "device_id" : 1, "partition_ids" : [], "hostname" : "mc-gb-nvl-020-001-switch", "model" : "N5110_LD", "health" : "Healthy" }, ... ] }

    • Status: 404 - Not Found

      • Description: The domain with one of specified UUIDs could not be found

      • Content

        Copy
        Copied!
                    

        { "error": "the domain with the uuid '4b47fa0f-7dbc-474b-b521-fc3ffc6dc7da' could not be found" }

Get Ports

  • Description: Gets a list of ports

  • URL:GET /ufmRestV2/plugin/nmxaggr/v1/inventory/ports

  • Parameters:

    Parameter

    Type

    Description

    Required

    Example

    Possible Values

    domain_uuid

    string

    Return only ports belonging to the domain with this UUID

    false

    ?domain_uuid=2e469d41-3102-485c-aa8b-58ba3c97f6bb

    domain_uuids

    list of strings

    Return only ports belonging to domains with UUIDs provided in a comma separated list

    false

    ?domain_uuids=2e469d41-3102-485c-aa8b-58ba3c97f6bb,d7c35619-fa03-4bb7-bce2-e4e37614b38b

  • Responses:

    • Status: 200 - OK

      • Description: List of ports is successfully returned

      • Content:

        Copy
        Copied!
                    

        { "ports" : [ { "domain_name" : "locujoko", "domain_uuid" : "4b47fa0f-7dbc-474b-b521-fc3ffc6dc7df", "domain_controller_address" : {"host" : "10.245.16.133", "port" : 6901}, "last_updated" : "2025-05-21T17:12:14+03:00", "port_uid" : "02e7-7ac6-7738-9e59", "device_uid" : "0954-1c38-eb6e-07c0", "peer_device_uid" : "0000-0002-2f8a-e90c", "system_port_num" : 0, "compute_port_num" : 4, "subnet_prefix" : "0000-0000-0000-0000", "base_lid" : 2123, "type" : "GPU", "physical_state" : "Link Up", "logical_state" : "Active", "port_num" : 5, "peer_port_num" : 1, "cage_num" : 0, "cage_split_num" : 0, "cage_port_num" : 0, "rail" : 0, "plane" : 0, "link_rate" : 0, "partition_ids" : [32766], "is_sdn_port" : false, "contain_and_drain" : false }, ... ] }

    • Status: 404 - Not Found

      • Description: The domain with one of specified UUIDs could not be found

      • Content:

        Copy
        Copied!
                    

        { "error": "the domain with the uuid '4b47fa0f-7dbc-474b-b521-fc3ffc6dc7da' could not be found" }

  • Description: Gets a list of links

  • URL: GET /ufmRestV2/plugin/nmxaggr/v1/inventory/links

  • Parameters:

    Parameter

    Type

    Description

    Required

    Example

    Possible Values

    domain_uuid

    string

    Return only links belonging to the domain with this UUID

    false

    ?domain_uuid=2e469d41-3102-485c-aa8b-58ba3c97f6bb

    domain_uuids

    list of strings

    Return only links belonging to domains with UUIDs provided in a comma separated list

    false

    ?domain_uuids=2e469d41-3102-485c-aa8b-58ba3c97f6bb,d7c35619-fa03-4bb7-bce2-e4e37614b38b

  • Responses:

    • Status: 200 - OK

      • Description: List of links is successfully returned

      • Content

        Copy
        Copied!
                    

        { "links" : [ { "domain_name" : "locujoko", "domain_uuid" : "4b47fa0f-7dbc-474b-b521-fc3ffc6dc7df", "domain_controller_address" : {"host" : "10.245.16.133", "port" : 6901}, "last_updated" : "2025-05-21T17:12:14+03:00", "endpoint_a" : { "location" : {"chassis_id" : 1, "slot_id" : 1, "host_id" : 1}, "device_uid" : "0d32-6108-ad29-dd4c", "device_id" : 4, "port_num" : 12, "cage_num" : 2147483647, "cage_split_num" : 2147483647, "cage_port_num" : 2147483647 }, "endpoint_b" : { "location" : {"chassis_id" : 1, "slot_id" : 10, "host_id" : 1}, "device_uid" : "0000-0002-381b-8cd3", "device_id" : 2, "port_num" : 1, "cage_num" : 2147483647, "cage_split_num" : 2147483647, "cage_port_num" : 2147483647 }, "type" : "GPU", "state" : "Active" }, ... ] }

    • Status: 404 - Not Found

      • Description: The domain with one of specified UUIDs could not be found

      • Content:

        Copy
        Copied!
                    

        { "error": "the domain with the uuid '4b47fa0f-7dbc-474b-b521-fc3ffc6dc7da' could not be found" }

Get Partitions

  • Description: Get a list of partitions

  • URL: GET /ufmRestV2/plugin/nmxaggr/v1/allocations/partitions

  • Parameters:

    Parameter

    Type

    Description

    Required

    Example

    Possible Values

    domain_uuid

    string

    Return only partitions belonging to the domain with this UUID

    false

    ?domain_uuid=2e469d41-3102-485c-aa8b-58ba3c97f6bb

    domain_uuids

    list of strings

    Return only partitions belonging to domains with UUIDs provided in a comma separated list

    false

    ?domain_uuids=2e469d41-3102-485c-aa8b-58ba3c97f6bb,d7c35619-fa03-4bb7-bce2-e4e37614b38b

    partition_id

    string (hex)

    Return only partitions with this ID

    false

    0x2e

    health

    string

    Filter returned partitions by a value of the "health" property

    false

    ?health=unhealthy

    healthy, degraded, unhealthy

  • Responses:

    • Status: 200 - OK

      • Description: List of partitions is successfully returned

      • Content:

        Copy
        Copied!
                    

        { "partitions": [ {       "domain_name": "nevexitu", "domain_uuid": "01ffdf80-5362-42ac-8191-289460699ea8", "partition_id": "0x7ffe", "name": "Default Partition", "type": "UID Based", "num_gpus": 36, "health": "Healthy", "resiliency_mode": "Undefined", "num_allocated_multicast_groups": 0, "multicast_groups_limit": 0, "last_updated": "2025-10-22T15:26:05+03:00"     }, . . . ] }

    • Status: 404 - Not Found

      • Description: The domain with one of specified UUIDs could not be found

      • Content:

        Copy
        Copied!
                    

        { "error": "the domain with the uuid '4b47fa0f-7dbc-474b-b521-fc3ffc6dc7da' could not be found" }

Create Partition

  • Description: Create a new partition

  • URL: POST /ufmRestV2/plugin/nmxaggr/v1/allocations/partitions

  • Request Body:

    Parameter

    Type

    Description

    Required

    Possible values

    Default Value

    domain_uuid

    string

    UUID of a domain in which a new partition will be created

    true

    partition_id

    string (hex)

    Must be unique in a domain

    true

    The valid range of values is from "0x1" to "0x7ffd". The value "0x7ffe" is reserved for the default partition

    name

    string

    If not provided, it will be generated from the partiion_id

    false

    For the partition_id "0xe1" the generated name will be "prt_0xe1"

    type

    string

    true

    • Location Based - A location-based partition is bound to the resources in the assigned location. The location is static, but the resources in the location are dynamic. This implies that if a compute tray that belongs to a partition is moved to a different slot, the partition loses the GPUs that migrated with the compute tray

    • UID Based - A UID-based partition is bound to the GPU UIDs even when the UIDs move between different locations in a domain

    compute_node_locations

    array

    Array of location objects for compute nodes to be allocated to the created partition. All GPUs belonging to specified compute nodes will be allocated to the partition. If not provided, an empty partition will be created

    false

    location.chassis_id

    integer

    true

    location.slot_id

    integer

    true

    Copy
    Copied!
                

    { "domain_uuid":"a260d0ea-c2a0-499d-8367-2656034f6439", "partition_id":"0xe1", "type":"Location Based", "compute_node_locations": [ {"chassis_id":1,"slot_id":3}, {"chassis_id":1,"slot_id":4} ] }

  • Responses:

    • Status: 200 - OK

      • Description: A partition was successfully created

Delete Partition

  • Description: Delete an existing partition. Deleting a partition is the release of associated partition resources such as GPUs, a unique name, multicast reservations, and the partition ID

  • URL: DELETE /ufmRestV2/plugin/nmxaggr/v1/allocations/partitions

  • Request Body:

    Parameter

    Type

    Description

    Required

    Possible values

    Default Value

    domain_uuid

    string

    UUID of a domain in which a new partition will be created

    true

    partition_id

    string (hex)

    Must be unique in a domain

    true

    The valid range of values is from "0x1" to "0x7ffd". The value "0x7ffe" is reserved for the default partition

    Copy
    Copied!
                

    { "domain_uuid":"a260d0ea-c2a0-499d-8367-2656034f6439", "partition_id":"0xe1", }

  • Responses:

    • Status: 200 - OK

      • Description: A partition was successfully deleted

Update Partition

  • Description: Update a partition (add/remove compute nodes to/from a partition)

  • URL: PUT /ufmRestV2/plugin/nmxaggr/v1/allocations/partitions

  • Request Body:

    Parameter

    Type

    Description

    Required

    Possible values

    Default Value

    domain_uuid

    string

    UUID of a domain in which a new partition will be created

    true

    partition_id

    string (hex)

    Must be unique in a domain

    true

    The valid range of values is from "0x1" to "0x7ffd". The value "0x7ffe" is reserved for the default partition

    compute_node_locations

    array

    Array of location objects for compute nodes that will be allocated to the partition after update. All GPUs belonging to specified compute nodes will be allocated to the partition. All compute nodes allocated to the partition and not mentioned in this parameter will be removed from the partition

    true

    location.chassis_id

    integer

    true

    location.slot_id

    integer

    true

    Copy
    Copied!
                

    { "domain_uuid":"a260d0ea-c2a0-499d-8367-2656034f6439", "partition_id":"0xe1", "compute_node_locations": [ {"chassis_id":1,"slot_id":4}, {"chassis_id":1,"slot_id":5} ] }

  • Responses:

    • Status: 200 - OK

      • Description: A partition was successfully updated

Get Allocatable Compute Nodes

  • Description: Get a list of allocatable compute nodes

  • URL: GET /ufmRestV2/plugin/nmxaggr/v1/allocations/compute_nodes

  • Parameters:

    Parameter

    Type

    Description

    Required

    Example

    Possible Values

    domain_uuid

    string

    Return only compute nodes belonging to the domain with this UUID

    false

    ?domain_uuid=2e469d41-3102-485c-aa8b-58ba3c97f6bb

    domain_uuids

    list of strings

    Return only compute nodes belonging to domains with UUIDs provided in a comma separated list

    false

    ?domain_uuids=2e469d41-3102-485c-aa8b-58ba3c97f6bb,d7c35619-fa03-4bb7-bce2-e4e37614b38b

    partition_id

    string (hex)

    Must be unique in a domain

    false

    ?partition_id=0xfe

    The valid range of values is from "0x1" to "0x7ffd". The value "0x7ffe" is reserved for the default partition

    is_allocated

    boolean

    If true, allocated compute nodes will be returned, otherwise free compute nodes will be returned

    false

    ?is_allocated=true

  • Responses:

    • Status: 200 - OK

      • Description: List of allocatable compute nodes is successfully returned

      • Content:

        Copy
        Copied!
                    

        { "compute_nodes": [ {       "domain_name": "lokiboli", "domain_uuid": "a260d0ea-c2a0-499d-8367-2656034f6439", "partition_id": "0xfe", "name": "lokiboli.chas-01.slot-04", "location": { "chassis_id": 1, "slot_id": 4 }, "last_updated": "2025-10-24T16:58:20+03:00"     }, . . . ] }

    • Status: 404 - Not Found

      • Description: The domain with one of specified UUIDs could not be found

      • Content:

        Copy
        Copied!
                    

        { "error": "the domain with the uuid '4b47fa0f-7dbc-474b-b521-fc3ffc6dc7da' could not be found" }

Get Allocatable GPUs

  • Description: Get a list of allocatable GPUs

  • URL: GET /ufmRestV2/plugin/nmxaggr/v1/allocations/gpus

  • Parameters:

    Parameter

    Type

    Description

    Required

    Example

    Possible Values

    domain_uuid

    string

    Return only compute nodes belonging to the domain with this UUID

    false

    ?domain_uuid=2e469d41-3102-485c-aa8b-58ba3c97f6bb

    domain_uuids

    list of strings

    Return only compute nodes belonging to domains with UUIDs provided in a comma separated list

    false

    ?domain_uuids=2e469d41-3102-485c-aa8b-58ba3c97f6bb,d7c35619-fa03-4bb7-bce2-e4e37614b38b

    partition_id

    string (hex)

    Must be unique in a domain

    false

    ?partition_id=0xfe

    The valid range of values is from "0x1" to "0x7ffd". The value "0x7ffe" is reserved for the default partition

    is_allocated

    boolean

    If true, allocated compute nodes will be returned, otherwise free compute nodes will be returned

    false

    ?is_allocated=true

  • Responses:

    • Status: 200 - OK

      • Description: List of allocatable GPUs is successfully returned

      • Content:

        Copy
        Copied!
                    

        { "compute_nodes": [ {       "domain_name": "nevexitu", "domain_uuid": "01ffdf80-5362-42ac-8191-289460699ea8", "partition_id": "0x7ffe", "name": "nevexitu.chas-01.slot-18.gpu-03", "device_uid": "0dbe-676c-44ee-0809", "device_location": { "chassis_id": 1, "slot_id": 18, "host_id": 1, "device_id": 3 }, "last_updated": "2025-10-24T17:02:23+03:00"     }, . . . ] }

    • Status: 404 - Not Found

      • Description: The domain with one of specified UUIDs could not be found

      • Content:

        Copy
        Copied!
                    

        { "error": "the domain with the uuid '4b47fa0f-7dbc-474b-b521-fc3ffc6dc7da' could not be found" }

Get Domains Allocations Stats

  • Description: Get a list of domains' allocations statistics

  • URL: GET /ufmRestV2/plugin/nmxaggr/v1/allocations/domains

  • Parameters:

    Parameter

    Type

    Description

    Required

    Example

    Possible Values

    domain_uuid

    string

    Return only compute nodes belonging to the domain with this UUID

    false

    ?domain_uuid=2e469d41-3102-485c-aa8b-58ba3c97f6bb

    domain_uuids

    list of strings

    Return only compute nodes belonging to domains with UUIDs provided in a comma separated list

    false

    ?domain_uuids=2e469d41-3102-485c-aa8b-58ba3c97f6bb,d7c35619-fa03-4bb7-bce2-e4e37614b38b

  • Responses:

    • Status: 200 - OK

      • Description: List of domains' allocations statistics is successfully returned

      • Content:

        Copy
        Copied!
                    

        { "domains": [ {       "domain_name": "lokiboli", "domain_uuid": "a260d0ea-c2a0-499d-8367-2656034f6439", "num_partitions": 1, "compute_node_allocations": { "total": 9, "allocated": 2, "free": 7 }, "gpu_allocations": { "total": 36, "allocated": 8, "free": 28 }, "last_updated": "2025-10-24T17:09:21+03:00"     }, . . . ] }

    • Status: 404 - Not Found

      • Description: The domain with one of specified UUIDs could not be found

      • Content:

        Copy
        Copied!
                    

        { "error": "the domain with the uuid '4b47fa0f-7dbc-474b-b521-fc3ffc6dc7da' could not be found" }

Get Stats

  • Description: Get statistics about inventory elements

  • URL: GET /ufmRestV2/plugin/nmxaggr/v1/stats

  • Responses:

    • Status: 200 - OK

      • Description: Statistics are successfully returned

      • Content:

        Copy
        Copied!
                    

        {   "stats": { "domains": { "total": 5, "healthy": 3, "degraded": 2, "unhealthy": 0, "unreachable": 0 }, "compute_nodes": { "total": 45, "healthy": 43, "degraded": 0, "unhealthy": 2 }, "switch_nodes": { "total": 45, "healthy": 45, "degraded": 0, "unhealthy": 0 }, "gpus": { "total": 180, "healthy": 178, "degraded": 0, "unhealthy": 2 }, "switches": { "total": 90, "healthy": 90, "degraded": 0, "unhealthy": 0 }, "partitions": { "total": 5, "healthy": 5, "degraded": 0, "unhealthy": 0 }, "compute_node_allocations": { "total": 45, "allocated": 38, "free": 7 }, "gpu_allocations": { "total": 180, "allocated": 152, "free": 28 }, "last_updated": "2025-10-24T17:10:53+03:00" }

© Copyright 2025, NVIDIA. Last updated on Nov 9, 2025