DHCP Redundancy

[Nathan McDavit-Van Fleet]

> Be warned that only HALF of the address space is "automatically" taken
> over by the surviving DHCP process (assuming the failover team consists
> of 2 servers) :
> "If one server fails, the other server will continue to renew leases
> out of the pool, and will allocate new addresses out of the roughly
> half of available addresses that it had when communications with the
> other server were lost. "

Is this true? Doesn't this make the fail-over rather limited then? I imagine
on a high churn network it wouldn't take long before the server is no longer
able to supply enough IP addresses to clients (having only legacy clients on
the other half). I wish we had the address space to have a 2x address-client
ratio but we don't.

Is there a way, in the case where a server is failed/off to enable the other
as just a regular functional server? 

Also I was surprised to come in this Monday and find that my server was
still rejecting some bind updates from the peer; because the "incoming
update was less critical than the outgoing one." I thought this was part of
the results of a failover? But since the services have not stopped for over
72 hours I imagine that this sort of mis match would have been ironed out by
now? 

Nathan Van Fleet

> -----Original Message-----
> From: dhcp-users-bounces+nmcdavit=alcor.concordia.ca at lists.isc.org
> [mailto:dhcp-users-bounces+nmcdavit=alcor.concordia.ca at lists.isc.org]
> On Behalf Of the Radio
> Sent: Sunday, November 28, 2010 5:49 AM
> To: dhcp-users at lists.isc.org
> Subject: Re: DHCP Redundancy
> 
> "If one fails, the other can take over and manage the full address
> space."
> 
> Be warned that only HALF of the address space is "automatically" taken
> over by the surviving DHCP process (assuming the failover team consists
> of 2 servers) :
> "If one server fails, the other server will continue to renew leases
> out of the pool, and will allocate new addresses out of the roughly
> half of available addresses that it had when communications with the
> other server were lost. "
> 
> v4.2.0-P1 option "auto-partner-down" does not work for now, then
> imposing to MANUALLY edit the leases file to set the surviving server
> into "PARTNER-DOWN" state (see dhcpd.conf help file).
> 
> Therefore, the failover efficiency depends greatly on the clients
> requests rate : I had a big wireless infrastructure that, whenever a
> controller shuts down, half of the 1000 APs request their address back
> - until you have a more than 2000 addresses subnet, some APs do not get
> their IP.
> 
> My main advice concerning failover (until partner-down state is really
> automatized) : use subnets large as twice as needed (if you have 200
> expected clients, set for a /23 subnet at least).
>