Feature Request - Use most specific available Prefix network size for discovered IP Addresses Mask

[Christophoclese]

Currently when doing Network Discovery and specifying a Prefix as a target, discovered IPs within that Prefix will either have a /32 mask or the same as the Prefix being scanned. This means you are not really able to scan an entire site (e.g. 10.28.0.0/16) as the results will be incorrect (/16) or not useful (/32).

It would be preferable that the mask from the most specific available Prefix is used. For example given the defined Prefixes:

10.28.0.0/16
10.28.24.0/22

The IP address 10.28.24.50 should be entered into NetBox as 10.28.24.50/22.

It seems like the current workaround would be to define Network Discovery targets only for the furthest leaf prefixes which creates additional complexity on the config side of things.

[leoparente]

hi @Christophoclese can you share more information? what is the prefix added for 10.28.24.50?
Because what you described in the issue is the current expected behavior (runner.go#L68). For instance, for a policy:

....
policies:
  network_discovery:
    discovery_2:
      scope:
        targets: [192.168.31.1/24, 192.168.31.56/28]
        fast_mode: True

I have the following result in NetBox:

[Christophoclese]

Hey @leoparente, thanks for the response. The behavior as you have described is also what I experience. Imagine you have these prefixes defined in NetBox:

• 10.28.0.0/16
  • 10.28.0.0/24
  • 10.28.24.0/22

This could continue with many child prefixes of 10.28.0.0/16. Instead of an Orb configuration that requires all child prefixes within the 10.28.0.0/16 space it would be nice to be able to specify one target of 10.28.0.0/16 and leverage any existing prefix information in NetBox to determine the mask. If an IP would fit into multiple NetBox Prefixes, use the most specific one available.