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

    WAN Acting Up With Multiple PF VMs

    Scheduled Pinned Locked Moved Virtualization
    4 Posts 3 Posters 1.4k 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.
    • V
      velvir
      last edited by

      Hey guys, we currently have 3 pfSense VMs on 2 different esxi 5.5 machines. PF1 is our main installation (2.2.3, 3 WAN Connections + 14 VLANs). PF2 and PF3 are PF 2.3 VMs that we want to replace our main one with (HA + CARP, 3 WANS + 14 VLANS). PF2 and PF3 have all the same settings as PF1 except DHCP is turned off and the WAN IPs are completely different. We're having an issue that whenever PF2 and PF3 are connected to WAN2 for an extended period of time, PF1 WAN2 is flooded with DHCP requests from VLAN1. My question is why is DHCP traffic appearing on the WAN side at all? All the WAN interfaces are statically assigned completely different IPs (including CARP VIPs).

      Current Setup:

      esxi host 1:
      pfSense 1 (2.2.3)
            WAN1: x.x.x.202
            WAN2: x.x.x.98
            WAN3: x.x.x.108
            VLAN1: 10.0.0.1
      pfSense 3 (2.3 HA+CARP, Backup)
            WAN1: x.x.x.203 (VIP: x.x.x.204)
            WAN2: x.x.x.99 (VIP: x.x.x.100)
            WAN3: x.x.x.109 (VIP: x.x.x.110)
            VLAN1: 10.0.0.5 (VIP: 10.0.0.6)

      esxi host 2:
      pfSense 2 (2.3 HA+CARP, Master)
            WAN1: x.x.x.205 (VIP: x.x.x.204)
            WAN2: x.x.x.101 (VIP: x.x.x.100)
            WAN3: x.x.x.111 (VIP: x.x.x.110)
            VLAN1: 10.0.0.2 (VIP: 10.0.0.6)

      VLAN1 DHCP and DNS is handled by Windows 2012 Domain Controller

      HA+CARP Sync Interface is an ethernet cable going from esxi host 1 to host 2 (no switch in between, only VM on said interface)
      Promiscuous mode, MAC Changes, and Forged Transmits enabled on all interfaces and vswitches
      Block Bogon and Local Traffic is enabled on all WAN interfaces

      Packet Capture of WAN2 on PF1 shows 10.0.1.94(Random Android Client) sending DHCP requests to 10.0.1.3 (Domain/DHCP server). When this happens packet loss jumps to ~70%
      We currently have PF2 and PF3 shutdown and WAN2 has not been acting up since.

      Thanks in advanced for any help  :)

      Edit: Forgot to mention there's only one client attached to PF2 and PF3 as a gateway (me, testing purposes) so there's not much, if any, traffic going through it.

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

        Do you have conflicting VHIDs between them? Those must be unique across pairs, if one pair has VHID 1, a different pair cannot use VHID 1. Otherwise you have a MAC address conflict. That shouldn't result in DHCP requests showing up on the wrong network, but I guess it's possible in theory that a duplicate MAC could screw up something on the network in such a way that unexpected things happen.

        1 Reply Last reply Reply Quote 0
        • V
          velvir
          last edited by

          The first time I set it up I accidentally set up two IPs on VHID 1 but I realized my mistake and changed it but it's still acting up even after fixing it  :(

          1 Reply Last reply Reply Quote 0
          • M
            magpiper
            last edited by

            FreeBSD 10 is NOT certified/supported on ESXi 5.1

            Refer to:

            https://forum.pfsense.org/index.php?topic=113220.0

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