Help with DHCP failover configuration

Phusion phusion2k at
Mon Aug 8 15:02:05 UTC 2011

I am in need of some assistance with DHCP failover. First of all both
of the servers are running 3.1-ESV. I've tried configured the servers
to run in a cooperating partner relationship to try and provide both
redundancy and load-balancing. As a test, I stopped the DHCP service
on the secondary server and eventually started to get messages like
this one on the primary server.

Aug  8 07:06:59 dhcp01 dhcpd: DHCPDISCOVER from xx:xx:xx:xx:xx:xx
(workstation) via fxp0: peer holds all free leases

Here is a copy of the dhcpd.master configuration file on each of the
two servers.

option domain-name "Test.local";
option domain-name-servers dns1.Test.local, dns2.Test.local;
ddns-update-style none;
default-lease-time 3600;
log-facility local7;
max-lease-time 3600;
min-lease-time 3600;
one-lease-per-client true;

subnet netmask {
    option routers;
    option netbios-name-servers dns1.Test.local, dns2.Test.local;
    option netbios-node-type 2;
    pool {
        failover peer "location1";

Here is a copy of the dhcpd.conf configuration file from the primary server.

failover peer "location1" {
    address dhcp01.Test.local;
    port 519;
    peer address dhcp02.Test.local;
    peer port 520;
    max-response-delay 60;
    max-unacked-updates 10;
    mclt 600;
    split 128;
    load balance max seconds 3;

include "/usr/local/etc/dhcpd.master";

Here is a copy of the dhcpd.conf configuration file from the secondary server.

failover peer "location1" {
    address dhcp02.Test.local;
    port 520;
    peer address dhcp01.Test.local;
    peer port 519;
    max-response-delay 60;
    load balance max seconds 3;

include "/usr/local/etc/dhcpd.master";

There are about 50 DHCP machines so I configured the pool to be 100
leases figuring that would be enough. Please advise.


More information about the dhcp-users mailing list