Server Fault is a question and answer site for system and network administrators. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

I'm trying to migrate from a situation with static ip addresses and a dhcp server on a router to one new dhcp server.

My idea was to add the current static ip addresses as static ip addresses in the dhcp server, so I can still manage them from that server, and to move the current dhcp leases to static ones as well, as these ip addresses are more or less used as static ones too, being referred to in hosts.allow etc on some machines for example.

I do however also still need dynamic addresses, and I'd like to know if I can for example define that my pool of dynamic addresses is from to, but have some exlusions in that range for the static addresses.

What would my dhcpd.conf have to look like to achieve that, can I just define the pool and define seperate static addresses and will the dhcpd take those out of the pool automatically?

I can't test this setup as this is a live network with a working dhcp on a router already, that's why I have to ask it up front.

share|improve this question
up vote 0 down vote accepted

I think you're getting your terminology confused sorry, let me try to clarify this for you.

Static addresses are just that, absolutely static, they're set on the individual host machines and have nothing to do with DHCP at all - as such they're not reliant on DHCP existing and don't have leases at all.

Fully dynamic addresses are given out from a DHCP server as part of a scope, do have leases and are succeptable to the various problems that DHCP brings such as what happens when the DHCP server falls over, what happens if/when someone switches on another DHCP server etc.

DHCP Reservations are almost identical to fully dynamic addresses with the exception that specified hosts always get the same IP address but are no different in the face of the problems listed above.

So getting to your question, yes you can define a dynamic scope and reserved IPs on one server but you have to manually define both sets and yes this does mean that you could misconfigure the server by having overlapped reserved addresses and dynamic scoped addresses - basically just don't do that by being careful and using something like a spreadsheet to keep you records and plans up to date.

share|improve this answer
Ok, this helps a lot! Too bad that the scope of dynamic addresses and reserved addresses can't overlap.. – datadevil Jan 10 '12 at 12:13
It's far from too bad, the point is you don't want to hand out already-reserved IP addressed to dynamic clients, otherwise you'd have two machines trying to use the same IP. What you can do is, as an example, have a dynamic scope of say then put your rserved IPs at the top of that rang (starting at .254 and moving down) then reducing the dynamic scope to end before the reserved IPs start. Does that make sense? – Chopper3 Jan 10 '12 at 12:27
yeah it makes sense, but I'd like it to be flexible that my pool is like you say e.g., but upon parsing the config file and finding reservations it'd take those out of the pool automatically, thus avoiding the double ip address problem. – datadevil Jan 11 '12 at 12:49
I suppose you could script something like that with lots of greps and seds but it'd be fragile as hell. – Chopper3 Jan 11 '12 at 13:32
yeah, not going in that direction – datadevil Jan 17 '12 at 15:46

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.