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

    How seamless is HA Failover in Pfsense

    Scheduled Pinned Locked Moved General pfSense Questions
    5 Posts 2 Posters 686 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.
    • U
      UserCo
      last edited by

      Hello,

      I have 2 identical pfSense Boxes setup in my lab and the failover works quite well. if I disconnect one interface on the primary it correctly fails over to the secondary. However I don't know how seamlessly the transition should be. If a download is running on a client behind the 2 firewalls and the primary fails, shouldn't then the download on the client continue without disconnecting as the failover should transfer all the states? because it does not in my setup. Am I missing something? when I try a ping test to lets say the google DNS, a failover only loses one to two pings.

      Any help is appreciated
      Thanks

      1 Reply Last reply Reply Quote 0
      • stephenw10S
        stephenw10 Netgate Administrator
        last edited by

        Yes, normally a TCP connection like that should continue. The states are sync'd between the nodes.

        U 1 Reply Last reply Reply Quote 0
        • U
          UserCo @stephenw10
          last edited by

          Thanks @stephenw10. I have setup all of the interfaces with carp VIPs, including the WAN, and have also adapted the outbound NAT config according to the docs. However, when I test it on a client, the download stops and does not even resume itself. If I look into the pftop, I see that the states are correctly copied over. What could be the issue here?

          1 Reply Last reply Reply Quote 0
          • stephenw10S
            stephenw10 Netgate Administrator
            last edited by

            The CARP VIPs must all failover. The states must be via the CARP VIPs so they are valid on both nodes. The interfaces must line up between the nodes so the states are valid.

            https://docs.netgate.com/pfsense/en/latest/troubleshooting/high-availability.html

            Steve

            U 1 Reply Last reply Reply Quote 2
            • U
              UserCo @stephenw10
              last edited by

              Thank you @stephenw10

              Turned out to be a config issue on my site which I missed. (forgot to add the sync states checkbox on the secondary node)

              1 Reply Last reply Reply Quote 2
              • First post
                Last post
              Copyright 2025 Rubicon Communications LLC (Netgate). All rights reserved.