Topics

1

17.7 Legacy Series / Flapping WAN interface

« on: September 29, 2017, 05:18:26 pm »

One of my WAN interfaces keep flapping every <10 seconds, I simply cannot figure out why. I've already attempted to move it to a different physical port on the firewall, no change.

This is a sample of the logs:

Sep 29 14:45:32   opnsense: /usr/local/etc/rc.newwanip: Switching default gateway to EM1_WAN_VDSL_PPPOE (OTHER-WAN-IP)
Sep 29 14:45:31   opnsense: /usr/local/etc/rc.newwanip: On (IP address: FLAPPING-WAN-IP) (interface: EM0_WAN_VIRGIN[wan]) (real interface: em0).
Sep 29 14:45:31   opnsense: /usr/local/etc/rc.newwanip: IP renewal is starting on 'em0'
Sep 29 14:45:31   kernel: arpresolve: can't allocate llinfo for FLAPPING-WAN-GW on em0
Sep 29 14:45:31   kernel: arpresolve: can't allocate llinfo for FLAPPING-WAN-GW on em0
Sep 29 14:45:31   kernel: arpresolve: can't allocate llinfo for FLAPPING-WAN-GW on em0
Sep 29 14:45:31   kernel: arpresolve: can't allocate llinfo for FLAPPING-WAN-GW on em0
Sep 29 14:45:31   kernel: arpresolve: can't allocate llinfo for FLAPPING-WAN-GW on em0
Sep 29 14:45:30   kernel: arpresolve: can't allocate llinfo for FLAPPING-WAN-GW on em0
Sep 29 14:45:30   kernel: em0: link state changed to UP
Sep 29 14:45:30   kernel: arpresolve: can't allocate llinfo for FLAPPING-WAN-GW on em0
Sep 29 14:45:30   kernel: arpresolve: can't allocate llinfo for FLAPPING-WAN-GW on em0
Sep 29 14:45:30   kernel: arpresolve: can't allocate llinfo for FLAPPING-WAN-GW on em0
Sep 29 14:45:29   kernel: arpresolve: can't allocate llinfo for FLAPPING-WAN-GW on em0
Sep 29 14:45:29   kernel: arpresolve: can't allocate llinfo for FLAPPING-WAN-GW on em0
Sep 29 14:45:28   kernel: arpresolve: can't allocate llinfo for FLAPPING-WAN-GW on em0
Sep 29 14:45:28   kernel: arpresolve: can't allocate llinfo for FLAPPING-WAN-GW on em0
Sep 29 14:45:28   kernel: em0: link state changed to DOWN
Sep 29 14:45:28   opnsense: /usr/local/etc/rc.linkup: HOTPLUG: Configuring interface wan
Sep 29 14:45:28   opnsense: /usr/local/etc/rc.linkup: DEVD Ethernet attached event for wan
Sep 29 14:45:28   configd.py: [79bc7ce3-a14c-4667-a1dc-df305fbe7fd4] Linkup starting em0
Sep 29 14:45:28   opnsense: /usr/local/etc/rc.linkup: Clearing states to old gateway FLAPPING-WAN-GW.
Sep 29 14:45:28   opnsense: /usr/local/etc/rc.linkup: DEVD Ethernet detached event for wan
Sep 29 14:45:28   configd.py: [6de0e49b-8cd3-4134-ae61-3ea0b8cdd10c] Linkup stopping em0

This repeats every just under 10 seconds. I'm not sure what to try now...

2

17.7 Legacy Series / WAN over PPPoE loses default IPv4 route in multi-WAN setup

« on: September 13, 2017, 04:55:24 pm »

I think this may be related to this bug, though I may well be wrong!

I can provide more logs, config etc as required later.

Setup: latest OPNSense installed on an AOpen i45GMt-HR with an HP NC364T NIC (so all NICs are Intel based); LAN is on the onboard NIC (em4), the two WANs (em0, em1) and a DMZ interface (em3) are on the NC364T.

First WAN (em0) is configured with IPv4 DHCP only; it's a cable ISP, there is no IPv6 connectivity provided.
Second WAN (em1) is configured with PPPoE, IPv4 DHCP and IPv6 DHCPv6 (PPPoE dual-stack). I also get an IPv6 DHCP-PD from this, works fine on the LAN etc. This is a VDSL2 ISP, and these gateways are set as the defaults.

I have a couple of gateway groups configured, but no rules in place that use them. I may try removing them to see if it makes a difference.

If I disable/re-enable the PPPoE connection or interface or IPv4 gateway, I completely lose the default IPv4 route unless I completely disable the PPPoE interface or IPv4 gateway.

It doesn't seem to matter what else I do, I just cannot seem to get the PPPoE default IPv4 route back, short of a reboot. The only thing that works is to disable the PPPoE WAN or PPPoE IPv4 gateway completely, which re-instates the Cable WAN as the default IPv4 route. If I don't do that, there is NO IPv4 default route.

I'm happy to provide logs, config details etc, just kindly let me know what would be required, and what tests would be useful. Thank you!