Utilization in prefix when IP range is defined

[haxorof]

I was looking into using IP ranges to define where you have DHCP inside you prefixes. When you define a IP range, which becomes a child range of a prefix, then the utilization changes even though you have not created a single IP within the IP range or the prefix. Is this how it should be? I thought it would show you 0% utilization if you have no IP addresses defined within that prefix.

See example from demo site below where I just created an IP range within an existing prefix:

As seen above the utilization here is at 57% even though no IP addresses has been created.

[candlerb]

Yes, I think that's correct. You've assigned a range of IPs to a DHCP server, which means those IPs are no longer available for any other purpose.

Of course, the number of DHCP clients which are active on your network at any time could be anywhere between 0% and 100% of that pool. But Netbox doesn't care about that; the point is, you've assigned all the addresses to the pool. You'd need to query your DHCP server(s) to determine the number of active leases, if that's what you care about. You might also want to set up monitoring, to alert when it gets close to full. That's a reflection of the dynamic usage of your network, not the static configuration of your network.

You wouldn't normally create IP addresses within a DHCP pool in Netbox, except when you're making static assignments for specific clients from within the dynamic pool range, and pushing those static assignments out to your DHCP server config. That shouldn't affect the overall utilization of the pool (i.e. Netbox shouldn't double-count the IP address).

[haxorof]

Thanks @candlerb!
Totally make sense when you explain it like that. In my case we do also track the DHCP addresses and DHCP reservations in NetBox so the the utilization then started me to wonder when I defined an IP range. 😊

Cheers!

[nogaff]

@candlerb Your answer assumes an IP range would only be used in the context of a DHCP pool, where the utilisation is unknown at any given moment in time and must therefore be assumed to be 100%.

But what if I want to document that a certain range is reserved for certain types of devices with static IP addresses, and only a small percentage of that range has actually been allocated so far?

Additionally, if you view the range itself, it will correctly show < 100% utilisation, while the prefix containing that range will still consider it to have 100% utilisation, so the utilisation calculation isn't even consistent between IP ranges and prefixes.

That seems like a bug to me.

However, if that was fixed then it would mean DHCP pools defined as an IP range would always show 0% utilisation, so there would also need to be a way to mark a range as a DHCP pool to force it to have 100% utilisation.

[DanSheps]

But what if I want to document that a certain range is reserved for certain types of devices with static IP addresses, and only a small percentage of that range has actually been allocated so far?

Then you should be using individual IP's if you don't want to see the allocation and not ranges. Or you can see the range utilization instead to gauge how much free space you have in a range.

If it bothers you, perhaps submit a FR to adjust the calculation based on the range status (Active == Calculate utilization, Reserved/Deprecated == Don't Calculate) or add a toggle such as "Is Pool" or similar.

[nogaff]

@DanSheps That's my point though. I am using individual IPs, but I can't document the ranges they fall into, without breaking the utilisation display on my prefixes, which just seems very odd to me.

I mean, if I've got a /24 prefix and the first 100 addresses are reserved for static IP addresses, with the remainder being for DHCP, but I've only connected 30 static devices, then clearly my utilisation is not 100% as the prefix utilisation would imply. It should be 73%, made up of 154 DHCP addresses, plus the 30 static devices (and the .0 and .255 at the beginning and end of the prefix).

Anyway, yes, I was indeed considering submitting an FR for an "Is pool" toggle as you suggest.

[candlerb]

Can we call it something more obvious please, like "treat as fully utilized".

("Is pool" already exists in other places in Netbox, and is used for different non-obvious reasons)

[JonathonReinhart]

Perhaps this is a separate issue or discussion, but one thing that concerns me here is that IP address which are part of a range still show up in the IP address list as "available".

For example:

• Prefix: 192.168.1.0/24
• Range: 192.168.1.51/24 - 192.168.1.225/24 (for DHCP)\

The Range does show up in the "Child Ranges" tab:

And it is properly accounted for on the main "Prefix" tab:

• Range IP addresses are not counted as "available"
• Range IP addresses are accounted for in utilization

⚠️ However, the IP addresses do appear here as "available":

Nothing is preventing someone from assigning an IP address in that range, and not realizing it's reserved for DHCP pool.

[aldemaroc]

Looks like its still an issued. Even tho IP blocks are marked as "Fully used", we still see all IPs as "Available".

[collinscochran]

Yeah, I'd love to see those appear as unavailable when looking at the prefix view.

[jronnblom]

I also voted for this suggestion. They shouldn't be visible in the Available IPs.

[Sprinterfreak]

Yes. This is exactly the point. Addresses claimed by a range must not be listed as usable within the prefix. This f** up things already.

I personally fixed this by not using ranges at all for now. I instead create all addresses with status DHCP or Reserved. I find ranges kind of obsolet because they are totally redundant to prefixes and addresses and therefore cause lots of confusion which repeatedly lead to duplicate address assignments.

[MarcHagen]

Hooking in here #9763 is somewhat related.
Specific this comment: #9763 (comment)

It seems something in the templating need to be changed to reflect this. It needs some good thinking to change.

netbox/netbox/ipam/tables/ip.py

Lines 50 to 58 in 28aee9b

	IPADDRESS_LINK = """
	{% if record.pk %}
	<a href="{{ record.get_absolute_url }}" id="ipaddress_{{ record.pk }}">{{ record.address }}</a>
	{% elif perms.ipam.add_ipaddress %}
	<a href="{% url 'ipam:ipaddress_add' %}?address={{ record.1 }}{% if object.vrf %}&vrf={{ object.vrf.pk }}{% endif %}{% if object.tenant %}&tenant={{ object.tenant.pk }}{% endif %}" class="btn btn-sm btn-success">{% if record.0 <= 65536 %}{{ record.0 }}{% else %}Many{% endif %} IP{{ record.0|pluralize }} available</a>
	{% else %}
	{% if record.0 <= 65536 %}{{ record.0 }}{% else %}Many{% endif %} IP{{ record.0|pluralize }} available
	{% endif %}
	"""

[telmich]

I was just running into the same issue, expecting a range to cover/show that an IP address is being used already, would be very natural to me.