DHCP failover setup with several relay agents

[John Wobus]

Hm, my The DHCP Handbook 2nd Edition has this paragraph on p160, 
Chapter 10:

> Failover Protocol Overview
>
> The DHCP synchronization protocol is called the _failover_protocol_ 
> because it was
> initially intended to provide a way for one DHCP server to act as a 
> primary server
> and for a second DHCP server to act as a backup.  In a very basic 
> failover configura-
> tion, the secondary server does not provide DHCP service when it is in 
> contact
> the primary; it simply accepts updates from the primary.  The 
> secondary will start to
> provide DHCP service only if it loses contact with the primary.  Thus, 
> DHCP service
> will _fail_over_ from the primary to the secondary server.  In a more 
> advanced configu-
> ration, both the primary and secondary servers provide service at the 
> same t ime,
> using a well-defined load-balancing algorithm to determine which 
> server answers
> which requests.

This paragraph left me thinking there might be some way to do 
non-load-balanced
failover using the protocol, but I don't recall ever run across 
anything about such a setup
in the book or man pages and I've concluded the paragraph might just be 
speaking generally
about server redundancy rather than the specific capabilities of the 
ISC server or the DHCP failover
protocol as spec'd at this point.  I assume the ISC server doesn't 
support this, from the answers
on this list and from the fact that the ISC developers haven't chimed 
in.

Failover could be done using server clustering/ha techniques, or by 
copying a lease file
from one server to another really often (with imperfect lease 
management, but practically
speaking, it could be a lot better than nothing). Re forwarders, 
whatever you could
get to work using such methods, you would have to use separate DHCP 
server addresses,
thus still have forwarders send to two addresses, or you would have to 
have two servers on the
same subnet so that an address could be moved from one server to the 
other, or you would
have to adjust routing tables, e.g., a host route.  Or you would have 
to make your forwarders
adapt at the time of failover.

John Wobus