Netgate Discussion Forum
    • Categories
    • Recent
    • Tags
    • Popular
    • Users
    • Search
    • Register
    • Login

    CARP sync renders both DHCPs enabled

    Scheduled Pinned Locked Moved HA/CARP/VIPs
    8 Posts 4 Posters 2.5k Views
    Loading More Posts
    • Oldest to Newest
    • Newest to Oldest
    • Most Votes
    Reply
    • Reply as topic
    Log in to reply
    This topic has been deleted. Only users with topic management privileges can see it.
    • M Offline
      marama
      last edited by

      2.3.1-RELEASE-p5 (amd64), on ESXi

      I'm using limiters, so I had to disable "Synchronize states".

      I've checked to sync "DHCP Server settings", and it does sync, but the problem is CARP enables DHCP on both servers, and I'm not sure that's the good idea. I thought the correct way would be only for master to have enabled DHCP. Is slave aware of the DHCP leases handed out by master so he hands out different IPs?

      I'm worried that even if I enter "static" gateway and DNS in DHCP, I might still get some conflicts if both DHCPs are enabled.

      TIA

      1 Reply Last reply Reply Quote 0
      • V Offline
        viragomann
        last edited by

        @marama:

        Is slave aware of the DHCP leases handed out by master so he hands out different IPs?

        Yes, if you have configured the option "Failover peer IP" at the master DHCP server. If you've done this the 2 boxes work as a "Failover Group". This shouldn't depend on state syncing.
        Check that in Status > DHCP Leases.

        1 Reply Last reply Reply Quote 0
        • C Offline
          cmb
          last edited by

          That's the expected and desired result, they're configured as a failover group. Not a problem, that's what you want.

          1 Reply Last reply Reply Quote 0
          • M Offline
            marama
            last edited by

            @viragomann:

            @marama:

            Is slave aware of the DHCP leases handed out by master so he hands out different IPs?

            Yes, if you have configured the option "Failover peer IP" at the master DHCP server. If you've done this the 2 boxes work as a "Failover Group". This shouldn't depend on state syncing.
            Check that in Status > DHCP Leases.

            You're right, "Failover peer IP" is currently disabled (on master and slave). What confuses me is this:
            https://doc.pfsense.org/index.php/Configuring_pfSense_Hardware_Redundancy_(CARP)

            "Enter the IP address of the secondary node in Failover peer IP. This will be automatically adjusted during synchronization."

            It wasn't filled during synchrinization in my case I believe.
            Anyway, I've filled in the LAN IP of the slave, now the slave has the LAN IP of the master in the same box, so I guess it's working fine.
            DHCP status on slave show this though:

            Pool Status
            Failover Group My State Since Peer State Since
            dhcp_lan (LAN) communications-interrupted 2016/07/25 17:58:03 normal 2016/07/25 17:57:32

            so I'm not sure if "communications-interrupte" sounds safe ;(

            1 Reply Last reply Reply Quote 0
            • V Offline
              viragomann
              last edited by

              "communications-interrupted" is shown if the DHCP server is disabled on master.

              1 Reply Last reply Reply Quote 0
              • M Offline
                marama
                last edited by

                @viragomann:

                "communications-interrupted" is shown if the DHCP server is disabled on master.

                Hmm, how can I debug that?
                I have DHCP on 5 interfaces, and on slave 4 of them are in state "recover", and the LAN one is in "communications-interrupted".
                On master, all 5 DHCP interfaces are "enabled".
                Any idea what could be wrong here?
                TIA

                1 Reply Last reply Reply Quote 0
                • M Offline
                  marama
                  last edited by

                  @marama:

                  @viragomann:

                  "communications-interrupted" is shown if the DHCP server is disabled on master.

                  Hmm, how can I debug that?
                  I have DHCP on 5 interfaces, and on slave 4 of them are in state "recover", and the LAN one is in "communications-interrupted".
                  On master, all 5 DHCP interfaces are "enabled".
                  Any idea what could be wrong here?
                  TIA

                  OK, this is the log output:

                  
                  2016-07-26T09:22:16+02:00 10.0.0.1 dhcpd: For info, please visit https://www.isc.org/software/dhcp/
                  2016-07-26T09:22:16+02:00 10.0.0.1 dhcpd: /etc/dhcpd.conf line 37: secondary may not define load balance settings.
                  2016-07-26T09:22:16+02:00 10.0.0.1 dhcpd:   split 128;
                  2016-07-26T09:22:16+02:00 10.0.0.1 dhcpd:          ^
                  2016-07-26T09:22:16+02:00 10.0.0.1 dhcpd: /etc/dhcpd.conf line 51: secondary may not define load balance settings.
                  2016-07-26T09:22:16+02:00 10.0.0.1 dhcpd:   split 128;
                  2016-07-26T09:22:16+02:00 10.0.0.1 dhcpd:          ^
                  2016-07-26T09:22:16+02:00 10.0.0.1 dhcpd: /etc/dhcpd.conf line 65: secondary may not define load balance settings.
                  2016-07-26T09:22:16+02:00 10.0.0.1 dhcpd:   split 128;
                  2016-07-26T09:22:16+02:00 10.0.0.1 dhcpd:          ^
                  2016-07-26T09:22:16+02:00 10.0.0.1 dhcpd: /etc/dhcpd.conf line 79: secondary may not define load balance settings.
                  2016-07-26T09:22:16+02:00 10.0.0.1 dhcpd:   split 128;
                  2016-07-26T09:22:16+02:00 10.0.0.1 dhcpd:          ^
                  2016-07-26T09:22:16+02:00 10.0.0.1 dhcpd: Configuration file errors encountered -- exiting
                  2016-07-26T09:22:16+02:00 10.0.0.1 dhcpd:
                  2016-07-26T09:22:16+02:00 10.0.0.1 dhcpd: If you think you have received this message due to a bug rather
                  2016-07-26T09:22:16+02:00 10.0.0.1 dhcpd: than a configuration issue please read the section on submitting
                  2016-07-26T09:22:16+02:00 10.0.0.1 dhcpd: bugs on either our web page at www.isc.org or in the README file
                  2016-07-26T09:22:16+02:00 10.0.0.1 dhcpd: before submitting a bug.  These pages explain the proper
                  2016-07-26T09:22:16+02:00 10.0.0.1 dhcpd: process and the information we find helpful for debugging..
                  2016-07-26T09:22:16+02:00 10.0.0.1 dhcpd:
                  2016-07-26T09:22:16+02:00 10.0.0.1 dhcpd: exiting.
                  2016-07-26T07:22:16+02:00 10.0.0.1 php-fpm[48524]: /status_services.php: The command '/usr/local/sbin/dhcpd -user dhcpd -group _dhcp -chroot /var/dhcpd -cf /etc/dhcpd.conf -pf /var/run/dhcpd.pid vmx0 vmx2_vlan21 vmx2_vlan22 vmx2_vlan23 vmx2_vlan100' returned exit code '1', the output was 'Internet Systems Consortium DHCP Server 4.3.3-P1 Copyright 2004-2016 Internet Systems Consortium. All rights reserved. For info, please visit https://www.isc.org/software/dhcp/ /etc/dhcpd.conf line 37: secondary may not define load balance settings.   split 128;          ^ /etc/dhcpd.conf line 51: secondary may not define load balance settings.   split 128;          ^ /etc/dhcpd.conf line 65: secondary may not define load balance settings.   split 128;          ^ /etc/dhcpd.conf line 79: secondary may not define load balance settings.   split 128;          ^ Configuration file errors encountered -- exiting  If you think you have received this message due to a bug rather than a configuration issue please read the section on submitting bugs
                  
                  

                  with:

                  
                  ...
                  failover peer "dhcp_lan" {
                    primary;
                    address 172.16.1.16;
                    port 519;
                    peer address 172.16.0.3;
                    peer port 520;
                    max-response-delay 10;
                    max-unacked-updates 10;
                    split 128;
                    mclt 600;
                  
                    load balance max seconds 3;
                  }
                  ...
                  
                  

                  so how do I get rid of the "split 128"?

                  1 Reply Last reply Reply Quote 0
                  • I Offline
                    iboulay
                    last edited by

                    To get rid of that split error, just get rid of that line in the Secondary config file.
                    It works for me ;D ;)

                    1 Reply Last reply Reply Quote 0
                    • First post
                      Last post
                    Copyright 2025 Rubicon Communications LLC (Netgate). All rights reserved.