Minowboard reboot bug



  • I have a remote network using a quad core dual ethernet minowboard from netgate. Every time I reboot my minnowboard the wan connection does not reconnect. I have to reboot my ont also just to get back online. Any advice is appreciated


  • Rebel Alliance Developer Netgate

    Anything in the main system log or DHCP log when that happens?



  • According to the dhcp log, it shows "connection closed" twice. Once before dhclient tries to connect and again 10 seconds after it finishes connecting


  • Rebel Alliance Developer Netgate

    Does it have the correct time on the clock / in the logs when it reboots?

    Are you using a bridge of some kind on the firewall?

    Is the WAN MAC address changing at each boot maybe?



  • The clock is correct. no bridge but I do have a linksys wrt3200acm running dd-wrt configured as a ap/switch connected to the lan port. Also the ont acts as a bridge between fiber and copper. I don't see mac address in the logs so I doubt it is changing. I noticed the following 2 lines in the status log.

    rc.linkup: DEVD Ethernet detached event for wan

    and

    Gateway alarm: WANGW (Addr:192.168.10.1 Alarm:1 RTT:0.000ms RTTsd:0.000ms Loss:100%)

    This second one is odd as that is not my wan ip address or even close to it.

    Also, for the first one, no one had been unplugging the wan ethernet wire.

    I forgot to mention, I am running 2.4.4p3 but I had these issues for many versions before it, ever since I received the board. I also should mention, I had my ont upgraded/replaced a little while ago and it made no difference. I had the issue before and I have it now.



  • sorry, for some odd reason the second one did not show. I will re-post it

    Gateway alarm: WANGW (Addr:192.168.10.1 Alarm:1 RTT:0.000ms RTTsd:0.000ms Loss:100%)



  • One more thing I just noticed

    /rc.newwanip: The command '/usr/local/sbin/unbound -c /var/unbound/unbound.conf' returned exit code '1', the output was '[1581544166] unbound[54749:0] error: bind: address already in use [1581544166] unbound[54749:0] fatal error: could not open ports'

    and

    /rc.newwanip: The command '/usr/local/sbin/unbound -c /var/unbound/unbound.conf' returned exit code '1', the output was '[1581544166] unbound[65021:0] error: bind: address already in use [1581544166] unbound[65021:0] fatal error: could not open ports'

  • Rebel Alliance Developer Netgate

    Your modem is probably handing out an RFC1918 address when it's not in sync, and pfSense picks that up but doesn't switch back to the public address after it gets a link.

    In your WAN settings, there is a box to reject DHCP leases from a specific address. Add 192.168.10.1 in there and then see if you can reboot successfully.



  • I tried it but no luck. Please look into the unbound errors I just posted


  • Netgate Administrator

    That unbound error is common at boot if it tried to start twice for some reason. I've never seen it cause a problem.

    The fact your WAN is pulling a private IP from the modem is almost certainly the issue here.

    Steve


  • Rebel Alliance Developer Netgate

    With that address in the "Reject leases from..." box, do you still see that private address at boot time in the logs? If so, look in the DHCP logs on pfSense, maybe its reporting a different DHCP server. If so, put the private address you see in the logs there. It would be on lines labeled DHCPOFFER from x.x.x.x or DHCPACK from x.x.x.x.



  • I don't see that address but I do see something else. I get a lot of these errors:

    arpresolve: can't allocate llinfo for 108.6.172.1 on igb0

    I also have a few of these

    /rc.linkup: DEVD Ethernet detached event for wan

    and a few seconds later

    /rc.linkup: DEVD Ethernet attached event for wan

    and

    /rc.newwanip: rc.newwanip: Failed to update wan IP, restarting...

    /rc.newwanip: rc.newwanip called with empty interface.

    Is it possible I have a bad nic?

    edit: I just noticed this also:

    /rc.newwanip: rc.newwanip: Failed to update wan IP, restarting...



  • fyi this is on verizon fios so no modem.


  • Rebel Alliance Developer Netgate

    Something upstream is handing you 192.168.10.1 in the previous cases, so either you have something else on the WAN segment serving DHCP, or it's your FIOS CPE/ONT/<whatever they want to call it> doing it.


  • Rebel Alliance Developer Netgate

    @techpro2004 said in Minowboard reboot bug:

    Is it possible I have a bad nic?

    Not likely, but you could swap the WAN/LAN assignments and see if the behavior changes.



  • Well We both got it wrong. In System/Routing/Gateways, Somehow I had 2 gateways. WANGW which had a gateway and monitor of 192.168.10.1 and I had WAN_DHCP. I switched the default to WAN_DHCP and then deleted WANGW and now it appears to be working. I don't know how WANGW got in there but I will look into it. FYI, fios is ipv4 only Thanks.


  • Rebel Alliance Developer Netgate

    It's always in the last place you look, eh? :-)



  • Absolutely, every time. That is because once you find it you stop looking :)


  • Netgate Administrator

    Hard to argue with that! 😁



  • Now that we know it is not a hardware issue, maybe you should move this thread to a more appropriate area of the forums. Thanks.


  • Rebel Alliance Developer Netgate

    That's the joke


Log in to reply