Allow nested LAG interface assignments #3828

New Issue

adam · 2025-12-29T18:31:23+01:00

adam commented

2025-12-29 18:31:23 +01:00

Originally created by @Kimmax on GitHub (Jul 2, 2020).

Originally assigned to: @jeremystretch on GitHub.

Environment

Python version: 3.7.3
NetBox version: 2.8.4

Proposed Functionality

The current implementation makes the creation of a stacked LAG bonding impossible. Although a LAG parent can be selected during modification, on confirmation the following error is thrown: "Link Aggregation Group (LAG) interfaces cannot have a parent LAG interface." - This is the right behavior and shouldn't be changed.

I'd like to see another virtual interface type Linux bond, with an option to specify the configured "bonding mode". Possible modes with their descriptions can be found on https://www.kernel.org/doc/Documentation/networking/bonding.txt#:~:text=mode,-Specifies%20one, section 2., under parameter "mode".

Use Case

Stacking bonding interfaces might be uncommon for network devices, but can be done in Linux.
As an example, we use two bonds, each configured as 802.3ad, and tie those together using another bond in the "active-backup" mode to ensure fast and high-available network access over two switches.

Database Changes

Not familiar with the database model of NetBox, but I guess the new type would have to be reflected in the model.

External Dependencies

No external dependencies

This request was already made in #3991 and was closed, because it didn't conform with the provided issue template.

Originally created by @Kimmax on GitHub (Jul 2, 2020). Originally assigned to: @jeremystretch on GitHub. ### Environment * Python version: 3.7.3 * NetBox version: 2.8.4 ### Proposed Functionality The current implementation makes the creation of a stacked LAG bonding impossible. Although a LAG parent can be selected during modification, on confirmation the following error is thrown: "Link Aggregation Group (LAG) interfaces cannot have a parent LAG interface." - This is the right behavior and shouldn't be changed. I'd like to see another virtual interface type _Linux bond_, with an option to specify the configured "bonding mode". Possible modes with their descriptions can be found on https://www.kernel.org/doc/Documentation/networking/bonding.txt#:~:text=mode,-Specifies%20one, section 2., under parameter "mode". ### Use Case Stacking bonding interfaces might be uncommon for network devices, but can be done in Linux. As an example, we use two bonds, each configured as 802.3ad, and tie those together using another bond in the "active-backup" mode to ensure fast and high-available network access over two switches. ### Database Changes Not familiar with the database model of NetBox, but I guess the new type would have to be reflected in the model. ### External Dependencies No external dependencies This request was already made in #3991 and was closed, because it didn't conform with the provided issue template.

adam added the status: accepted type: feature labels 2025-12-29 18:31:23 +01:00

adam closed this issue

2025-12-29 18:31:23 +01:00

adam commented

2025-12-29 18:31:24 +01:00

@stale[bot] commented on GitHub (Jul 16, 2020):

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. NetBox is governed by a small group of core maintainers which means not all opened issues may receive direct feedback. Please see our contributing guide.

@stale[bot] commented on GitHub (Jul 16, 2020): This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. NetBox is governed by a small group of core maintainers which means not all opened issues may receive direct feedback. Please see our [contributing guide](https://github.com/netbox-community/netbox/blob/develop/CONTRIBUTING.md).

adam commented

2025-12-29 18:31:24 +01:00

@jeremystretch commented on GitHub (Jul 16, 2020):

I don't know that it makes sense to add a separate type for bonded interfaces. Maybe it would be sufficient to relax the constraint on the existing LAG type?

@jeremystretch commented on GitHub (Jul 16, 2020): I don't know that it makes sense to add a separate type for bonded interfaces. Maybe it would be sufficient to relax the constraint on the existing LAG type?