Feedback on Fiber/Cable Path Visualization

[De-Backer]

Christian Rose,
First of all, compliments on the progress; it’s already looking very promising.

Regarding the fiber/cable paths in the Global Site Map:

• Do you already have a clear vision of how these will be visualized?

• Alternatively, would it be helpful to have some real-life examples for reference? (e.g., HDPE ducts, fiber optic cable routes with window-cuts, or splice boxes).

I assume these cable paths will be linked to dcim/cables/ and that a single path will be able to contain multiple dcim/cables.

Kind regards,
De-Backer

[DenDanskeMine]

If you have some references of how you model your fiver plz share! It would be of great help!

I'm currently working on a splice function that allows you to splice different fiber strains together. It is quite a buggy mess rn, so it will take some time for a final release

And yes I plan on adding cable linking to the paths / fiber

[De-Backer]

At its core:

• Fiber: The individual optical fiber.
• Buffer tube (or loose tube): A tube containing multiple fibers in a water-blocking gel. When performing a window cut, one strips back to this level.
• The cable (jacket): The outer casing that houses multiple buffer tubes.

Outside the cable:

• Microduct: A small tube designed to hold a single cable.
• HDPE Duct: A larger high-density polyethylene pipe that accommodates multiple microducts.
• Bore (or Conduit): A drill hole or underground passage containing multiple HDPE ducts.
• Trajectory (or Route): The final overall path of the installation.

How to model a window cut in NetBox.

cabinet window cut (With optional breakout points; normally, two loose tubes are cut at these locations, resulting in four loose tube endpoints.)

Now for the couplings:
HDPE ducts and microducts can be joined using couplers (or muffs). These points are crucial for maintenance: in the event of cable damage, they provide easy access to retrieve and repair the cable since they are not deeply buried. (In contrast, a bore can be located 20 to 50 meters underground.)
At these locations, a manhole (or handhole) can be installed. This allows a microduct to switch to a different HDPE duct, connect to a device, or even—in specific cases—loop back into the same HDPE duct (a window cut deviating from the standard route).

To make this functional in NetBox, I propose the following:
Use the ID from the dcim_cable table. This ID represents the loose tube if a window cut is performed; otherwise, it represents the cable as a whole. You can store this ID in an array (similar to how the GPS coordinates are handled) so that multiple cables or loose tubes can follow the same route. Consequently, a single cable can be part of multiple trajectories, and a single trajectory can contain multiple cables.

I will ask for a trajectory sample to determine the best way to translate this into NetBox. Most people use QGIS, which offers various export possibilities.

Is everything clear? If you have any questions or comments, just let me know.

update: add img "Outside the cable"

[DenDanskeMine]

Damn! This is awesome!!

I will begin to work on a basic draft, and update you here for some feedback.

Much appreciated 🤘🏻

[De-Backer]

Since I’m not 100% satisfied with my previous explanation; it could be clearer. Attempt #2, an addition:

Each section of Bore (or Conduit) and HDPE Duct is assigned a number. Should this be entered as data in NetBox? I suspect so, because the Microducts come in different colors. It’s useful to know which cable is in which Microduct and HDPE Duct just to make sure we cut the right tube :)
I am therefore looking at how modules fit into a device, for example:

• Trajectory (or Route): as a 'device'.
• Bore (or Conduit): as a 'module' of the Trajectory.
• An HDPE Duct: as a 'module' of the Trajectory or Bore (or Conduit).
• A Microduct: as a 'module' of the Trajectory, Bore (or Conduit), or HDPE Duct.

And so on, down to the level of the cable or even the Buffer tube (or loose tube) in the case of a window cut.

This way, everything inside an HDPE Duct is just one click away, as the HDPE Duct is the parent of the Microduct, which in turn is the parent of the cable, and so forth.

Another issue I see now involves the IDs of Cables in NetBox. For example, if I delete a Cable with ID foo that is also in your Microduct/cable 'module', and later create a new Cable that happens to get the same ID foo to solve this, I suggest using tags. If a cable doesn't have the required tag, it shouldn't appear in the duct.

Unfortunately, I haven't been able to create any QGIS data yet to serve as an import example; that will require some more time.

[DenDanskeMine]

Haha, I'll be honest - I might need some help with this one!

Right now I'm still figuring out the scope and direction, since I'm not a fiber guy at all. The hardest part is not knowing how people actually model their fiber infrastructure in NetBox - everyone seems to do it differently.

Your breakdown is really helpful though. The hierarchy you proposed (Trajectory > Bore > HDPE Duct > Microduct > Cable/Buffer tube) makes a lot of sense, and the module-as-parent approach is an interesting way to think about it.

I've been poking around with some early prototypes (screenshots below), but it's still pretty rough. One idea I'm leaning towards is an auto device creator — essentially a wizard that lets users create a set of prebuilt generic devices (splice boxes, manholes, etc.) so there's a consistent baseline to work with, rather than expecting everyone to have already modeled things a specific way.
In teh current state if a device have rear ports, an automatic button apperes that says:

The tag-based cable linking idea is interesting, but wouldn't that break the native NetBox cable linking? Since netbox-map currently relies on dcim/cables/ relationships to trace paths, filtering by tag rather than ID might disconnect it from NetBox's own cable topology. Would love your thoughts on how you'd handle that.

I'll try to redo what i have made to align bteter with what you just wrote!

ANd no rush on the QGIS export, but whenever you have something I'd love to use it as a real-world test case

[De-Backer]

Another issue I see now involves the IDs of Cables in NetBox. For example, if I delete a Cable with ID foo that is also in your Microduct/cable 'module', and later create a new Cable that happens to get the same ID foo to solve this, I suggest using tags. If a cable doesn't have the required tag, it shouldn't appear in the duct.

You can scrap this statement; I just tested it, and as long as you don't import cables with an ID included, the ID will always increment by 1. So, no problem.

The tag-based cable linking idea is interesting, but wouldn't that break the native NetBox cable linking? Since netbox-map currently relies on dcim/cables/ relationships to trace paths, filtering by tag rather than ID might disconnect it from NetBox's own cable topology. Would love your thoughts on how you'd handle that.

Note: This is about a single cable segment, not the end-to-end trace.
Physically, a cable runs through multiple conduits via couplers.
Thus, the Cable/Buffer tube serves as the ID-holder for that specific trajectory segment.

As an example: cable #922 runs through #5021 and #4024, among others. We therefore map #922 to 5021 and 4024 by assigning the ID of #922 to both.

Some conduits are intentionally missing because this also happens in reality; it is better to have no information than incorrect information.

[De-Backer]

update
Add support for cable bundles see:
netbox-community/netbox#21636
It would be best to wait and see how this develops; we may be able to utilize it then.

Additionally, this plugin (https://github.com/CESNET/cesnet_service_path_plugin) provides geographic path visualization for circuits. You can upload the path geometry as a file. Supported formats include GeoJSON (.geojson, .json), KML (.kml), and KMZ (.kmz).

[DenDanskeMine]

I will wait on netbox-community/netbox#21636 to be released to main before doing any release / edits to what i have already made in my dev-env.

Please still come with ideas if you have any.