Netgate Discussion Forum
    • Categories
    • Recent
    • Tags
    • Popular
    • Users
    • Search
    • Register
    • Login
    Introducing Netgate Nexus: Multi-Instance Management at Your Fingertips.

    Question about device discovery across subnets

    Scheduled Pinned Locked Moved Firewalling
    11 Posts 6 Posters 373 Views 6 Watching
    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.
    • G Offline
      ghar36k
      last edited by

      I have a Netgate 4200 and have been noticing blocked traffic in my firewall logs for two Windows computers on different subnets. The traffic seems to be tied to the windows update sharing service. The subnets are on different switches + wireless APs connected to different ports on my network interfaces. I don't have Avahi installed or configured and have a firewall rule set to block all traffic between the two interfaces.

      I'm wondering how they're seeing each other as when I use network scanning tools like fing or WiFiman they're not leaping across my vlans and spotting other devices.

      luckman212L keyserK johnpozJ 3 Replies Last reply Reply Quote 0
      • luckman212L Online
        luckman212 LAYER 8 @ghar36k
        last edited by luckman212

        are you using and/or also blocking IPv6 traffic between the subnets?

        G 1 Reply Last reply Reply Quote 0
        • G Offline
          ghar36k @luckman212
          last edited by

          @luckman212 said in Question about device discovery across subnets:

          are you using and/or also blocking IPv6 traffic between the subnets?

          The toggle "Allow IPv6" is not enabled on my 4200. And I have a generic firewall rule to block all IPv4/IPv6 traffic between those subnets.

          luckman212L 1 Reply Last reply Reply Quote 0
          • luckman212L Online
            luckman212 LAYER 8 @ghar36k
            last edited by luckman212

            I don't know all the technical details of how the Windows Update sharing or bandwidth-saving feature works. But, I could easily imagine it having a discovery method that works something like this:

            • PC1 on VLAN1 "phones home" to a coordination server at Microsoft, reporting some basic details about its network: LAN IP, WAN IP, MAC address of gateway etc
            • PC2 on VLAN2 phones home to same coordination server
            • Coordination server sees that PC1 and PC2 are likely behind the same NAT (since they share the same public IP, gateway MAC etc). It then instructs the 2 PCs to attempt to directly communicate via their LAN IPs

            Some careful wireshark/tcpdumping would probably reveal more, and possibly how to block this if you really want to. Seems like your block rule is already doing the right thing though.

            1 Reply Last reply Reply Quote 0
            • keyserK Online
              keyser Rebel Alliance @ghar36k
              last edited by

              @ghar36k Are your accesspoints using one SSID with authentication to assign which VLAN the clients land in?
              If thats the case, then remember you have to block multicast specifically in the Accesspoint, otherwise multicast to/from the clients will be distributed to both VLANs.

              Love the no fuss of using the official appliances :-)

              G 1 Reply Last reply Reply Quote 0
              • johnpozJ Offline
                johnpoz LAYER 8 Global Moderator @ghar36k
                last edited by johnpoz

                @ghar36k you prob want to look over this

                https://learn.microsoft.com/en-us/windows/deployment/do/delivery-optimization-configure

                Local area network (LAN)

                To define a peer group limited to your LAN, choose DODownloadMode (1), LAN-mode. This download mode setting includes any devices that share the same public IP address when they connect to the Internet (behind the same NAT) in a single peer group.

                Pretty sure @luckman212 got very close to how its trying to talk to devices on other networks behind your router.

                I would either just turn this off on the clients or your AD if running that these machines are a member of - or if you want allow the specific ports need between your subnets to allow for it.

                DO.jpg

                An intelligent man is sometimes forced to be drunk to spend time with his fools
                If you get confused: Listen to the Music Play
                Please don't Chat/PM me for help, unless mod related
                SG-4860 26.07 | Lab VMs 2.9.0, 26.07

                dennypageD 1 Reply Last reply Reply Quote 2
                • dennypageD Offline
                  dennypage @johnpoz
                  last edited by

                  @johnpoz said in Question about device discovery across subnets:

                  @ghar36k you prob want to look over this

                  https://learn.microsoft.com/en-us/windows/deployment/do/delivery-optimization-configure

                  And this: Microsoft: How Delivery Optimization works.

                  It uses Teredo, which has a STUN like capability.

                  1 Reply Last reply Reply Quote 1
                  • tinfoilmattT Offline
                    tinfoilmatt LAYER 8
                    last edited by tinfoilmatt

                    Just adding to the already-sound list of suggestions to check managed switches and AP's for: VLAN isolation via ACLs, and IGMP/MLD snooping/proxying.

                    block out log on { ix0 } inet from any to any
                    block out log on { ix0 } inet6 from any to any

                    G 1 Reply Last reply Reply Quote 0
                    • G Offline
                      ghar36k @tinfoilmatt
                      last edited by

                      @tinfoilmatt said in Question about device discovery across subnets:

                      Just adding to the already-sound list of suggestions to check managed switches and AP's for: VLAN isolation via ACLs, and IGMP/MLD snooping/proxying.

                      I have a managed UnFi switch and AP. Gateway mDNS Proxy is turned off which shouldn't forward multicast traffic across my VLANS. I have multicast filtering set to auto, which the documentation indicates will only forward multicast traffic within the VLAN and not across them. To troubleshoot a similar issue with an Apple device I has previously turned on Client Device Isolation for all of the WiFi networks/VLANs managed by the Ubiquiti switch/AP. I don't have IGMP snooping/proxying on from what I can see.

                      The only subnet that operates outside of my Unifi hardware is an old EERO device I have set to bridge mode. I can't set VLANs on that device so I'm wondering if there's some kind of segregation issue there. The EERO is the one that's connected to the separate port and is not connected to the managed switch on my Unifi hardware.

                      @keyser said in Question about device discovery across subnets:

                      @ghar36k Are your accesspoints using one SSID with authentication to assign which VLAN the clients land in?
                      If thats the case, then remember you have to block multicast specifically in the Accesspoint, otherwise multicast to/from the clients will be distributed to both VLANs.

                      I was thinking about doing this. It's not really a permanent issue as one of the Windows PCs is owned by a house guest. I'm more wondering if I've messed my configuration up somewhere that would let them see/speak to each other without the above mentioned methods. I did notice something similar happening with MDNS traffic between an iPhone and an Apple TV which is why I made this thread.

                      @luckman212 said in Question about device discovery across subnets:

                      I don't know all the technical details of how the Windows Update sharing or bandwidth-saving feature works. But, I could easily imagine it having a discovery method that works something like this:

                      • PC1 on VLAN1 "phones home" to a coordination server at Microsoft, reporting some basic details about its network: LAN IP, WAN IP, MAC address of gateway etc
                      • PC2 on VLAN2 phones home to same coordination server
                      • Coordination server sees that PC1 and PC2 are likely behind the same NAT (since they share the same public IP, gateway MAC etc). It then instructs the 2 PCs to attempt to directly communicate via their LAN IPs

                      Some careful wireshark/tcpdumping would probably reveal more, and possibly how to block this if you really want to. Seems like your block rule is already doing the right thing though.

                      This makes a lot of sense. The same sort of thing happened between two Apple devices too a few months ago (owned by different Apple accounts). If Apple has the same sort of system it explains how they're seeing each other.

                      1 Reply Last reply Reply Quote 1
                      • G Offline
                        ghar36k @keyser
                        last edited by

                        @keyser said in Question about device discovery across subnets:

                        @ghar36k Are your accesspoints using one SSID with authentication to assign which VLAN the clients land in?
                        If thats the case, then remember you have to block multicast specifically in the Accesspoint, otherwise multicast to/from the clients will be distributed to both VLANs.

                        They're all using separate SSIDs.

                        1 Reply Last reply Reply Quote 0
                        • G Offline
                          ghar36k
                          last edited by

                          Whoops! I quoted keyser in that big block of text above when I meant to quote part of the comment about wireshark/tcpdumping. It won't let me edit it anymore to fix it. Sorry for the misquote and excessive pings.

                          1 Reply Last reply Reply Quote 0
                          • First post
                            Last post
                          Copyright 2026 Rubicon Communications LLC (Netgate). All rights reserved.
                          Privacy Policy · Cookie Policy