Netbox Get Containerlab Inventory Task

task api name: get_containerlab_inventory

This task designed to provide Containerlab workers with inventory data sourced from Netbox to deploy lab topologies.

Get Containerlab Inventory Sample Usage

Below is an example of how to fetch Containerlab topology inventory data from Netbox for two devices named fceos4 and fceos5.

nf#netbox get containerlab-inventory devices fceos4 fceos5 lab-name foobar
--------------------------------------------- Job Events -----------------------------------------------
31-May-2025 13:10:14.477 7d434ed4e24c4a69af5d52797d7a187e job started
31-May-2025 13:10:14.525 INFO netbox-worker-1.1 running netbox.get_containerlab_inventory  - Fetching devices data from Netbox
31-May-2025 13:10:14.594 INFO netbox-worker-1.1 running netbox.get_containerlab_inventory  - Node added fceos4
31-May-2025 13:10:14.600 INFO netbox-worker-1.1 running netbox.get_containerlab_inventory  - Node added fceos5
31-May-2025 13:10:14.606 INFO netbox-worker-1.1 running netbox.get_containerlab_inventory  - Fetching connections data from Netbox
31-May-2025 13:10:15.211 INFO netbox-worker-1.1 running netbox.get_containerlab_inventory  - Link added fceos5:eth1 - fceos4:eth1
31-May-2025 13:10:15.217 INFO netbox-worker-1.1 running netbox.get_containerlab_inventory  - Link added fceos5:eth2 - fceos4:eth2
31-May-2025 13:10:15.225 INFO netbox-worker-1.1 running netbox.get_containerlab_inventory  - Link added fceos5:eth3 - fceos4:eth3
31-May-2025 13:10:15.232 INFO netbox-worker-1.1 running netbox.get_containerlab_inventory  - Link added fceos5:eth4 - fceos4:eth4
31-May-2025 13:10:15.238 INFO netbox-worker-1.1 running netbox.get_containerlab_inventory  - Link added fceos5:eth6 - fceos4:eth6
31-May-2025 13:10:15.244 INFO netbox-worker-1.1 running netbox.get_containerlab_inventory  - Link added fceos5:eth7 - fceos4:eth7
31-May-2025 13:10:15.250 INFO netbox-worker-1.1 running netbox.get_containerlab_inventory  - Link added fceos5:eth8 - fceos4:eth101
31-May-2025 13:10:15.257 INFO netbox-worker-1.1 running netbox.get_containerlab_inventory  - Link added fceos5:eth11 - fceos4:eth11
31-May-2025 13:10:15.580 INFO netbox-worker-1.1 running netbox.get_containerlab_inventory  - Renaming fceos4 interfaces
31-May-2025 13:10:15.587 INFO netbox-worker-1.1 running netbox.get_containerlab_inventory  - Renaming fceos5 interfaces
31-May-2025 13:10:15.808 7d434ed4e24c4a69af5d52797d7a187e job completed in 1.331 seconds

--------------------------------------------- Job Results --------------------------------------------

netbox-worker-1.1:
  mgmt:
    ipv4-subnet: 172.100.100.0/24
    network: br-foobar
  name: foobar
  topology:
    links:
    - endpoints:
      - interface: eth1
        node: fceos5
      - interface: eth1
        node: fceos4
      type: veth
    - endpoints:
      - interface: eth2
        node: fceos5
      - interface: eth2
        node: fceos4
      type: veth
    - endpoints:
      - interface: eth3
        node: fceos5
      - interface: eth3
        node: fceos4
      type: veth
    - endpoints:
      - interface: eth4
        node: fceos5
      - interface: eth4
        node: fceos4
      type: veth
    - endpoints:
      - interface: eth6
        node: fceos5
      - interface: eth6
        node: fceos4
      type: veth
    - endpoints:
      - interface: eth7
        node: fceos5
      - interface: eth7
        node: fceos4
      type: veth
    - endpoints:
      - interface: eth8
        node: fceos5
      - interface: eth101
        node: fceos4
      type: veth
    - endpoints:
      - interface: eth11
        node: fceos5
      - interface: eth11
        node: fceos4
      type: veth
    nodes:
      fceos4:
        image: ceosimage:4.30.0F
        kind: ceos
        mgmt-ipv4: 172.100.100.2
        ports:
        - 12000:22/tcp
        - 12001:23/tcp
        - 12002:80/tcp
        - 12003:161/udp
        - 12005:830/tcp
        - 12006:8080/tcp
      fceos5:
        image: ceosimage:4.30.0F
        kind: ceos
        mgmt-ipv4: 172.100.100.3
        ports:
        - 12007:22/tcp
        - 12008:23/tcp
        - 12009:80/tcp
        - 12010:161/udp
        - 12011:443/tcp
        - 12012:830/tcp
        - 12013:8080/tcp

nf#

NORFAB Netbox Get Containerlab Inventory Command Shell Reference

NorFab shell supports these command options for Netbox get_containerlab_inventory task:

nf#man tree netbox.get.containerlab-inventory
root
└── netbox:    Netbox service
    └── get:    Query data from Netbox
        └── containerlab-inventory:    Query Netbox and construct Containerlab inventory
            ├── timeout:    Job timeout
            ├── workers:    Filter worker to target, default 'any'
            ├── verbose-result:    Control output details, default 'False'
            ├── lab-name:    Lab name to generate lab inventory for
            ├── tenant:    Tenant name to generate lab inventory for
            │   ├── tenant:    Filter devices by tenants
            │   ├── device-name-contains:    Filter devices by name pattern
            │   ├── model:    Filter devices by models
            │   ├── platform:    Filter devices by platforms
            │   ├── region:    Filter devices by regions
            │   ├── role:    Filter devices by roles
            │   ├── site:    Filter devices by sites
            │   ├── status:    Filter devices by statuses
            │   └── tag:    Filter devices by tags
            ├── devices:    List of devices to generate lab inventory for
            ├── progress:    Display progress events, default 'True'
            ├── netbox-instance:    Name of Netbox instance to pull inventory from
            ├── ipv4-subnet:    IPv4 management subnet to use for lab, default '172.100.100.0/24'
            ├── image:    Docker image to use for all nodes
            └── ports:    Range of TCP/UDP ports to use for nodes, default '[12000, 13000]'
nf#

Python API Reference

Retrieve and construct Containerlab inventory from NetBox data.

Containerlab node details must be defined under device configuration context norfab.containerlab path, for example:

{
    "norfab": {
        "containerlab": {
            "kind": "ceos",
            "image": "ceos:latest",
            "mgmt-ipv4": "172.100.100.10/24",
            "ports": [
                {10000: 22},
                {10001: 830}
            ],

            ... any other node parameters ...

            "interfaces_rename": [
                {
                    "find": "eth",
                    "replace": "Eth",
                    "use_regex": false
                }
            ]
        }
    }
}

For complete list of parameters refer to Containerlab nodes definition.

Special handling given to these parameters:

• lab_name - if not provided uses tenant argument value as a lab name
• kind - uses device platform field value by default
• image - uses image value if provided, otherwise uses {kind}:latest
• interfaces_rename - a list of one or more interface renaming instructions, each item must have find and replace defined, optional use_regex flag specifies whether to use regex based pattern substitution.

To retrieve topology data from Netbox at least one of these arguments must be provided to identify a set of devices to include into Containerlab topology:

• tenant - topology constructed using all devices and links that belong to this tenant
• devices - creates topology only using devices in the lists
• filters - list of device filters to retrieve from Netbox and add to topology

If multiple of above arguments provided, resulting lab topology is a sum of all devices matched.

Parameters:

Name	Type	Description	Default
job	Job	NorFab Job object containing relevant metadata	required
lab_name	(str, Mandatory)	Name of containerlab to construct inventory for.	None
tenant	str	Construct topology using given tenant's devices	None
filters	list	List of filters to apply when retrieving devices from NetBox.	None
devices	list	List of specific devices to retrieve from NetBox.	None
instance	str	NetBox instance to use.	None
image	str	Default containerlab image to use,	None
ipv4_subnet	(str, Optional)	Management subnet to use to IP number nodes starting with 2nd IP in the subnet, in assumption that 1st IP is a default gateway.	'172.100.100.0/24'
ports	(tuple, Optional)	Ports range to use for nodes.	(12000, 15000)
ports_map	(dict, Optional)	dictionary keyed by node name with list of ports maps to use,	None
cache	Union[bool, str]	Cache usage options: True: Use data stored in cache if it is up to date, refresh it otherwise. False: Do not use cache and do not update cache. "refresh": Ignore data in cache and replace it with data fetched from Netbox. "force": Use data in cache without checking if it is up to date.	False

Returns:

Name	Type	Description
dict	Result	Containerlab inventory dictionary containing lab topology data

Source code in norfab\workers\netbox_worker\containerlab_inventory_tasks.py

@Task(fastapi={"methods": ["GET"], "schema": NetboxFastApiArgs.model_json_schema()})
def get_containerlab_inventory(
    self,
    job: Job,
    lab_name: str = None,
    tenant: Union[None, str] = None,
    filters: Union[None, list] = None,
    devices: Union[None, list] = None,
    instance: Union[None, str] = None,
    image: Union[None, str] = None,
    ipv4_subnet: str = "172.100.100.0/24",
    ports: tuple = (12000, 15000),
    ports_map: Union[None, dict] = None,
    cache: Union[bool, str] = False,
) -> Result:
    """
    Retrieve and construct Containerlab inventory from NetBox data.

    Containerlab node details must be defined under device configuration
    context `norfab.containerlab` path, for example:

    ```
    {
        "norfab": {
            "containerlab": {
                "kind": "ceos",
                "image": "ceos:latest",
                "mgmt-ipv4": "172.100.100.10/24",
                "ports": [
                    {10000: 22},
                    {10001: 830}
                ],

                ... any other node parameters ...

                "interfaces_rename": [
                    {
                        "find": "eth",
                        "replace": "Eth",
                        "use_regex": false
                    }
                ]
            }
        }
    }
    ```

    For complete list of parameters refer to
    [Containerlab nodes definition](https://containerlab.dev/manual/nodes/).

    Special handling given to these parameters:

    - `lab_name` - if not provided uses `tenant` argument value as a lab name
    - `kind` - uses device platform field value by default
    - `image` - uses `image` value if provided, otherwise uses `{kind}:latest`
    - `interfaces_rename` - a list of one or more interface renaming instructions,
        each item must have `find` and `replace` defined, optional `use_regex`
        flag specifies whether to use regex based pattern substitution.

    To retrieve topology data from Netbox at least one of these arguments must be provided
    to identify a set of devices to include into Containerlab topology:

    - `tenant` - topology constructed using all devices and links that belong to this tenant
    - `devices` - creates topology only using devices in the lists
    - `filters` - list of device filters to retrieve from Netbox and add to topology

    If multiple of above arguments provided, resulting lab topology is a sum of all
    devices matched.

    Args:
        job: NorFab Job object containing relevant metadata
        lab_name (str, Mandatory): Name of containerlab to construct inventory for.
        tenant (str, optional): Construct topology using given tenant's devices
        filters (list, optional): List of filters to apply when retrieving devices from NetBox.
        devices (list, optional): List of specific devices to retrieve from NetBox.
        instance (str, optional): NetBox instance to use.
        image (str, optional): Default containerlab image to use,
        ipv4_subnet (str, Optional): Management subnet to use to IP number nodes
            starting with 2nd IP in the subnet, in assumption that 1st IP is a default gateway.
        ports (tuple, Optional): Ports range to use for nodes.
        ports_map (dict, Optional): dictionary keyed by node name with list of ports maps to use,
        cache (Union[bool, str], optional): Cache usage options:

            - True: Use data stored in cache if it is up to date, refresh it otherwise.
            - False: Do not use cache and do not update cache.
            - "refresh": Ignore data in cache and replace it with data fetched from Netbox.
            - "force": Use data in cache without checking if it is up to date.

    Returns:
        dict: Containerlab inventory dictionary containing lab topology data
    """
    devices = devices or []
    filters = filters or []
    nodes, links = {}, []
    ports_map = ports_map or {}
    endpts_done = []  # to deduplicate links
    instance = instance or self.default_instance
    # handle lab name and tenant name with filters
    if lab_name is None and tenant:
        lab_name = tenant
    # add tenant filters
    if tenant:
        filters = filters or [{}]
        for filter in filters:
            filter["tenant__name"] = tenant

    # construct inventory
    inventory = {
        "name": lab_name,
        "topology": {"nodes": nodes, "links": links},
        "mgmt": {"ipv4-subnet": ipv4_subnet, "network": f"br-{lab_name}"},
    }
    ret = Result(
        task=f"{self.name}:get_containerlab_inventory",
        result=inventory,
        resources=[instance],
    )
    mgmt_net = ipaddress.ip_network(ipv4_subnet)
    available_ips = list(mgmt_net.hosts())[1:]

    # run checks
    if not available_ips:
        raise ValueError(f"Need IPs to allocate, but '{ipv4_subnet}' given")
    if ports:
        available_ports = list(range(ports[0], ports[1]))
    else:
        raise ValueError(f"Need ports to allocate, but '{ports}' given")

    # check Netbox status
    netbox_status = self.get_netbox_status(job=job, instance=instance)
    if netbox_status.result[instance]["status"] is False:
        ret.failed = True
        ret.messages = [f"Netbox status is no good: {netbox_status}"]
        return ret

    # retrieve devices data
    log.debug(
        f"Fetching devices from {instance} Netbox instance, devices '{devices}', filters '{filters}'"
    )
    job.event("fetching devices data from Netbox")
    nb_devices = self.get_devices(
        job=job,
        filters=filters,
        devices=devices,
        instance=instance,
        cache=cache,
    )

    # form Containerlab nodes inventory
    for device_name, device in nb_devices.result.items():
        node = device["config_context"].get("norfab", {}).get("containerlab", {})
        # populate node parameters
        if not node.get("kind"):
            if device["platform"]:
                node["kind"] = device["platform"]
            else:
                msg = (
                    f"{device_name} - has no 'kind' of 'platform' defined, skipping"
                )
                log.warning(msg)
                job.event(msg, severity="WARNING")
                continue
        if not node.get("image"):
            if image:
                node["image"] = image
            else:
                node["image"] = f"{node['kind']}:latest"
        if not node.get("mgmt-ipv4"):
            if available_ips:
                node["mgmt-ipv4"] = f"{available_ips.pop(0)}"
            else:
                raise RuntimeError("Run out of IP addresses to allocate")
        if not node.get("ports"):
            node["ports"] = []
            # use ports map
            if ports_map.get(device_name):
                node["ports"] = ports_map[device_name]
            # allocate next-available ports
            else:
                for port in [
                    "22/tcp",
                    "23/tcp",
                    "80/tcp",
                    "161/udp",
                    "443/tcp",
                    "830/tcp",
                    "8080/tcp",
                ]:
                    if available_ports:
                        node["ports"].append(f"{available_ports.pop(0)}:{port}")
                    else:
                        raise RuntimeError(
                            "Run out of TCP / UDP ports to allocate."
                        )

        # save node content
        nodes[device_name] = node
        job.event(f"node added {device_name}")

    # return if no nodes found for provided parameters
    if not nodes:
        msg = f"{self.name} - No devices found in Netbox"
        log.error(msg)
        ret.failed = True
        ret.messages = [
            f"{self.name} - No devices found in Netbox, "
            f"devices - '{devices}', filters - '{filters}'"
        ]
        ret.errors = [msg]
        return ret

    job.event("fetching connections data from Netbox")

    # query interface connections data from netbox
    nb_connections = self.get_connections(
        job=job, devices=list(nodes), instance=instance, cache=cache
    )
    # save connections data to links inventory
    while nb_connections.result:
        device, device_connections = nb_connections.result.popitem()
        for interface, connection in device_connections.items():
            # skip non ethernet links
            if connection.get("termination_type") != "interface":
                continue
            # skip orphaned links
            if not connection.get("remote_interface"):
                continue
            # skip connections to devices that are not part of lab
            if connection["remote_device"] not in nodes:
                continue
            endpoints = []
            link = {
                "type": "veth",
                "endpoints": endpoints,
            }
            # add A node
            endpoints.append(
                {
                    "node": device,
                    "interface": interface,
                }
            )
            # add B node
            endpoints.append({"node": connection["remote_device"]})
            if connection.get("breakout") is True:
                endpoints[-1]["interface"] = connection["remote_interface"][0]
            else:
                endpoints[-1]["interface"] = connection["remote_interface"]
            # save the link
            a_end = (
                endpoints[0]["node"],
                endpoints[0]["interface"],
            )
            b_end = (
                endpoints[1]["node"],
                endpoints[1]["interface"],
            )
            if a_end not in endpts_done and b_end not in endpts_done:
                endpts_done.append(a_end)
                endpts_done.append(b_end)
                links.append(link)
                job.event(
                    f"link added {endpoints[0]['node']}:{endpoints[0]['interface']}"
                    f" - {endpoints[1]['node']}:{endpoints[1]['interface']}"
                )

    # query circuits connections data from netbox
    nb_circuits = self.get_circuits(
        job=job, devices=list(nodes), instance=instance, cache=cache
    )
    # save circuits data to hosts' inventory
    while nb_circuits.result:
        device, device_circuits = nb_circuits.result.popitem()
        for cid, circuit in device_circuits.items():
            # skip circuits not connected to devices
            if not circuit.get("remote_interface"):
                continue
            # skip circuits to devices that are not part of lab
            if circuit["remote_device"] not in nodes:
                continue
            endpoints = []
            link = {
                "type": "veth",
                "endpoints": endpoints,
            }
            # add A node
            endpoints.append(
                {
                    "node": device,
                    "interface": circuit["interface"],
                }
            )
            # add B node
            endpoints.append(
                {
                    "node": circuit["remote_device"],
                    "interface": circuit["remote_interface"],
                }
            )
            # save the link
            a_end = (
                endpoints[0]["node"],
                endpoints[0]["interface"],
            )
            b_end = (
                endpoints[1]["node"],
                endpoints[1]["interface"],
            )
            if a_end not in endpts_done and b_end not in endpts_done:
                endpts_done.append(a_end)
                endpts_done.append(b_end)
                links.append(link)
                job.event(
                    f"link added {endpoints[0]['node']}:{endpoints[0]['interface']}"
                    f" - {endpoints[1]['node']}:{endpoints[1]['interface']}"
                )

    # rename links' interfaces
    for node_name, node_data in nodes.items():
        interfaces_rename = node_data.pop("interfaces_rename", [])
        if interfaces_rename:
            job.event(f"renaming {node_name} interfaces")
        for item in interfaces_rename:
            if not item.get("find") or not item.get("replace"):
                log.error(
                    f"{self.name} - interface rename need to have"
                    f" 'find' and 'replace' defined, skipping: {item}"
                )
                continue
            pattern = item["find"]
            replace = item["replace"]
            use_regex = item.get("use_regex", False)
            # go over links one by one and rename interfaces
            for link in links:
                for endpoint in link["endpoints"]:
                    if endpoint["node"] != node_name:
                        continue
                    if use_regex:
                        renamed = re.sub(
                            pattern,
                            replace,
                            endpoint["interface"],
                        )
                    else:
                        renamed = endpoint["interface"].replace(pattern, replace)
                    if endpoint["interface"] != renamed:
                        msg = f"{node_name} interface {endpoint['interface']} renamed to {renamed}"
                        log.debug(msg)
                        job.event(msg)
                        endpoint["interface"] = renamed

    return ret