split scoping

[Chris Buxton]

On Nov 24, 2009, at 1:11 PM, Simon Hobson wrote:
> Ausmus, Matt wrote:
>> What are the pros and cons to using just split scoping between a couple of ISC servers vs. using the draft fault tolerance mechanisms built into ISC DHCPd?  1 obvious con is that DHCP leases would be split between 2 different machines.  What are some others and are there any pros to this setup?
> 
> I assume split scoping means running two independent servers with non-overlapping address allocations.
> 
> Well the first issue is that without manual intervention (and problems when you come to recover), both servers must have enough address allocation to service the entire client base (excluding fixed addresses which can be common across multiple servers). So if you have 200 clients, you need over 400 available addresses - of which over half will be unused.
> 
> The next problem is that if a server fails under split scoping, then any clients using a lease from that server WILL get a new IP address.

The usual way to set up split scopes in the Microsoft world is to use one subnet, with disjoint address pools within (or the same address pool but half excluded, which is functionally the same thing).

With ISC DHCP, this has a problem: If the servers both consider themselves authoritative, they'll NAK each other's leases when they come up for renewal. But if you use different subnets, rather than splitting the subnet, then things like zeroconf won't work.

I'm sure someone will correct me if I'm mistaken.

Chris Buxton
Professional Services
Men & Mice