[ale] dhcpd misbehaving?

[Robert L. Harris]

I have set up 2 VM guests ( currently on the same hardware ) which is
configured as such:

Host 1:  172.27.0.11
  Bind 9 : Master
  dhcpd : Primary
  Postfix relay

Host 2: 172.27.0.12
  Bind 9 : Master
  dhcpd : Failover
  Postfix relay

Heartbeat Failover IP : 172.27.0.10

  I have all the components and and working happily.  I want bind 9
resolving on both hosts for the sake of load but dhcpd only responding on 1
IP and postfix is to only answer/send on the .10.  Failover tests have done
what I want, the way I want.

  The problem I'm running into is dhcpd will work great for 3-5 days and
then all of a sudden start throwing errors:
Mar 30 15:50:01 basvc01 dhcpd: DHCPDISCOVER from c8:9c:dc:da:b4:bf via
172.27.4.1: not responding (recovering)
Mar 30 15:50:01 basvc01 dhcpd: DHCPDISCOVER from 00:50:56:a0:57:aa via
172.27.6.1: not responding (recovering)
Mar 30 15:50:05 basvc01 dhcpd: DHCPDISCOVER from c8:9c:dc:da:b4:bf via
172.27.4.1: not responding (recovering)

If I disable the failover function and only run on host1, dhcpd runs great
and without error.

Here is the failover definition included in the main dhcpd.conf:
failover peer "failover-partner" {
    primary;
    # on Primary
    address 172.27.0.11;
    # on Secondary
    #address 172.27.0.12;
    port 519;
    # on Primary
    peer address 172.27.0.12;
    # on Secondary
    #peer address 172.27.0.11;
    peer port 520;
    max-response-delay 120;
    max-unacked-updates 30;
    mclt 7200;
    split 128;
    load balance max seconds 5;
}

Here is the head of one of my dhcpd config files ( including multiple for
multiple vlans in dhcpd.conf ).  Note I commented out the failover lines so
it'll work for now:

# SCM - VLAN 806
# 172.27.6.0/24 subnet
subnet 172.27.6.0 netmask 255.255.255.0 {
        pool {
#               failover peer "failover-partner";
                range   172.27.6.200 172.27.6.250;
        }
        option routers                          172.27.6.1;
        option subnet-mask                      255.255.255.0;
        option broadcast-address                172.27.6.255;
        option time-offset                      -25200; # MDT
        option ntp-servers                      time.nist.gov,
time-a.timefreq.bldrdoc.gov;
        use-host-decl-names                     on;
        one-lease-per-client                    on;
        default-lease-time                      21600;
        max-lease-time                          43200;
        option domain-name-servers              172.27.0.10;
        option domain-name                      "scm.baengr.company.com";
        option domain-search                    "scm.baengr.company.com",
"company.local";
        zone scm.baengr. {
                primary 172.27.0.11;
                secondary 172.27.0.12;
        }

        zone 6.27.172.in-addr.arpa. {
                primary 172.27.0.11;
                secondary 172.27.0.12;
        }
}
# Static Assignments


Any thoughts?  Everything I can find on google says it's a port timing or
firewall issue. These are 2 Linux VM's on the same Host, no firewall
between them, no iptables either currently.

Robert
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.ale.org/pipermail/ale/attachments/20150402/18aed55f/attachment.html>