WAN going DOWN and UP

[diablo]

Hi All,

OK bit of a long one perhaps, but I'll try to get as my details in here as possible in the hope you guys can advise:

I have an APU2c I bought from Appliance Shop EU, which has been running since July OPNsense. The LAN port is connected to a TP-Link gigabit switch, and the WAN is connected to an Arris cable modem in bridge mode.

The other day the internet started playing up, and while I thought it was my ISP, after digging around a little I realised it was actually the OPNsense box. I see the following in /var/log/system.log

Dec 18 06:19:38 gw kernel: igb1: link state changed to DOWN
Dec 18 06:19:43 gw kernel: igb1: link state changed to UP
Dec 18 06:19:43 gw opnsense: /usr/local/etc/rc.newwanip: rc.newwanip: Informational is starting igb1.
Dec 18 06:19:43 gw opnsense: /usr/local/etc/rc.newwanip: rc.newwanip: on (IP address: 84.236.XXX.XXX) (interface: WAN[wan]) (real interface: igb1).
Dec 18 06:19:44 gw opnsense: /usr/local/etc/rc.newwanip: ROUTING: setting IPv4 default route to 84.236.XXX.XXX
Dec 18 06:19:45 gw opnsense: /usr/local/etc/rc.linkup: ROUTING: setting IPv4 default route to 84.236.XXX.XXX
Dec 18 06:19:49 gw configd.py: [1f8b627c-9ef2-455d-8ebb-461fa6f493b8] updating dyndns wan
Dec 18 06:19:49 gw devd: Executing '/usr/local/opnsense/service/configd_ctl.py interface linkup stop igb1'
Dec 18 06:19:49 gw configd.py: [e777e332-6673-4d15-a526-6f50506858f7] Linkup stopping igb1
Dec 18 06:19:50 gw opnsense: /usr/local/etc/rc.linkup: DEVD Ethernet detached event for wan
Dec 18 06:19:51 gw opnsense: /usr/local/etc/rc.linkup: Clearing states to old gateway 84.236.XXX.XXX.
Dec 18 06:19:51 gw devd: Executing '/usr/local/opnsense/service/configd_ctl.py interface linkup start igb1'
Dec 18 06:19:51 gw configd.py: [d48abdfb-32cb-4d2d-9873-c1c607bea4c7] Linkup starting igb1
Dec 18 06:19:51 gw opnsense: /usr/local/etc/rc.linkup: DEVD Ethernet attached event for wan
Dec 18 06:19:51 gw opnsense: /usr/local/etc/rc.linkup: HOTPLUG: Configuring interface wan
Dec 18 06:19:52 gw kernel: igb1: link state changed to DOWN
Dec 18 06:19:56 gw kernel: igb1: link state changed to UP
Dec 18 06:19:57 gw opnsense: /usr/local/etc/rc.newwanip: rc.newwanip: Informational is starting igb1.
Dec 18 06:19:57 gw opnsense: /usr/local/etc/rc.newwanip: rc.newwanip: on (IP address: 84.236.XXX.XXX) (interface: WAN[wan]) (real interface: igb1).
Dec 18 06:19:57 gw opnsense: /usr/local/etc/rc.newwanip: ROUTING: setting IPv4 default route to 84.236.XXX.XXX
Dec 18 06:19:59 gw opnsense: /usr/local/etc/rc.linkup: ROUTING: setting IPv4 default route to 84.236.XXX.XXX
Dec 18 06:20:02 gw configd.py: [d5c8ac48-6fbb-42ef-bb28-1da414a1db63] updating dyndns wan
Dec 18 06:20:03 gw devd: Executing '/usr/local/opnsense/service/configd_ctl.py interface linkup stop igb1'
Dec 18 06:20:03 gw configd.py: [fde0dc91-0b40-4732-a84c-2e61174a7876] Linkup stopping igb1
Dec 18 06:20:03 gw opnsense: /usr/local/etc/rc.linkup: DEVD Ethernet detached event for wan
Dec 18 06:20:04 gw opnsense: /usr/local/etc/rc.linkup: Clearing states to old gateway 84.236.XXX.XXX.
Dec 18 06:20:04 gw devd: Executing '/usr/local/opnsense/service/configd_ctl.py interface linkup start igb1'
Dec 18 06:20:04 gw configd.py: [07f90642-9f5c-4717-b380-fa059fd61cd5] Linkup starting igb1
Dec 18 06:20:05 gw opnsense: /usr/local/etc/rc.linkup: DEVD Ethernet attached event for wan
Dec 18 06:20:05 gw opnsense: /usr/local/etc/rc.linkup: HOTPLUG: Configuring interface wan
Dec 18 06:20:05 gw kernel: igb1: link state changed to DOWN
Dec 18 06:20:10 gw kernel: igb1: link state changed to UP

This repeats constantly until a reboot. The odd thing is it can go anything from 5 minutes until 11 hours without failing, then suddenly again it begins.

To check it was not the ISP, I spoofed the OPNsense MAC address to a MacBook Pro and connected from the Arris directly to the MacBook Pro, and the connection is solid, no ping drops. If I then reattach the cable to the OPNsense it immediately continues sending the WAN interface UP and DOWN once again.

The only thing I do notice in the logs is a line like:

Dec 18 07:47:07 gw kernel: amdtemp0: <AMD CPU On-Die Thermal Sensors> on hostb5

I've tried swapping cables, rebooting the Arris, everything... I'm torn between if this is a HW or SW issue. I've looked at top output when it's been happening and not seen any problem processes.

So, your thoughts please guys?

All the best,

[diablo]

[diablo]

Well, I decided to purchase another power supply. When the problem was happening I swapped the supply for the new one, and now 24 hours later no issues... so fingers crossed that was the issue.

If not, I'll be back here.

Thanks to the #OPNsense channel and to Appliance Shop for their help on IRC.

All the best for Christmas and New Year.

[diablo]