Weird vLAN DHCP problem

I’m having a very odd problem with DHCP assignment. I have a Balance20X bonding three WANs: Starlink (via Ethernet WAN); LTE (via the in-built modem); and DSL (via a USB-Ethernet adapter connected to a CenturyLink modem in bridge mode). Wifi is provided by a UniFi system with four APs. I have four vLANs: primary, guest, IoT, and a “bypass”, which exists so that users can join it to bypass the SpeedFusion Cloud VPN (via an outbound rule tied to the relevant subnet). All are advertised as separate SSIDs, the Unifi APs tag the traffic, and then I have the vLANs set up in the 20x.

Everything worked perfectly until a few days ago, but now when any client joins either the Guest or Bypass vLANs, the router fails to issue a DHCP lease. The log instantly fills up with requests for every possible IP address in the range (I’m using a /24 subnet), saying that the IP address has been assigned to a device with MAC 00:00:00:00:00:00 and then finally throws an error that the DHCP pool is full. This happens even when there are no clients on that subnet and when there haven’t been any clients for days if not weeks.

Any new vLANs I create exhibit this same behavior, but the existing Primary and IoT vLANs work perfectly for some reason. The only thing I can think of is that those vLANs always have some clients, while the Guest and Bypass ones are usually empty (as, of course, is any new one). Still, very weird. I tried rebooting the router and get the same error.

I have not tried a factory reset, since I was able to work around the problem by using dnsmasq on a Linux box I had lying around doing $stuff to assign leases in those ranges and telling the 20X to use it as a relay. That solution works great, but I’m still puzzled as to what is going on (and why it started seemingly randomly; it’s not like I updated the firmware or anything). I did hard power cycle it a day before I noticed this problem, as the web UI became completely unresponsive (wouldn’t even load), so maybe that corrupted something?

In any event, I’m curious if anyone has any troubleshooting ideas or suggestions before I go down the factory reset path.

Thanks for such a great product. I’ve been incredibly happy with it; it’s turned three lousy rural ISPs into one good one! (I guess it’s not fair to call Starlink “lousy”, but I have a lot of obstructions, so at least in my use case, it’s unreliable.

@Tyler_Griffin ,

My I know do you created a ticket for the DHCP lease issue ? For this, support team may need to check from the device in-order to understand how the DHCP packets being forwarded.

If the ticket is not created, can you please create one here ?
https://ticket.peplink.com/ticket/new/public

Thank you

Old thread but this is a bug and was just fixed in a beta firmware 8.4.1s185-build5498 if anyone else is experiencing it . You might have to open a ticket to get the updated firmware

Thanks for following up! I’ll just wait for the release version since my dnsmasq workaround is functional for now, and it’s a critical system on which I’d rather not run a beta.

Still great to hear they took care of it; hope this helps anyone else who runs into the issue.