Can Aggregates Take into account IP Addresses not just Prefixes?

[65156]

As title.

I'd like to be able to see available IP addresses within some of my IP aggregates, for example if I want to monitor IP availability for some Public Subnets, I dont want to have to create /32 prefix's

Could there be an option within a prefix to select whether it reports by Prefix or IP utilization as obviously we cant report both.

[jeremystretch]

Reporting on individual IPs used is actually a lot simpler than accounting for child prefixes, since it's just a count of all IP addresses within the aggregate prefix as a fraction of the total available IPs. For example, if an aggregate 192.168.0.0/16 has 4827 IPs defined, its utilization is 4827/65536 (~7%).

It may be reasonable to display both prefix and IP utilization in the aggregate table, depending on how much room we need for the description.

[Gelob]

Maybe a dupe or similar to #40 ?

[candlerb]

Reporting on individual IPs used is actually a lot simpler than accounting for child prefixes

Not much use for IPv6 though! I suggest for that case, the figure of interest is the number of /64's touched.

if an aggregate 192.168.0.0/16 has 4827 IPs defined, its utilization is 4827/65536 (~7%)

Yes this is possibly an interesting figure, but may be misleading. It can hide that you have some prefixes at or close to 100% utilisation, and no way to grow them without serious renumbering.

I think the real issue is that Netbox only reports utilisation at the level of Aggregate, not at the Prefix level. If you could see utilisation of each prefix separately, that would go a long way I think.

Within a prefix:

1. If it contains only IP addresses, then you can report the utilisation as (IPs defined / total space)
2. If it contains prefixes, then you can report the utilisation as (child prefix space / total space)
3. If it contains both, then you count (child prefix space) + (IPs directly contained but not within child prefixes) - although that's admittedly an odd situation

And maybe (1) and (2) should be different colours.

[jeremystretch]

There doesn't seem to have been much interest in this so I'm closing it out.