4200 - LAN port flapping
-
I just configured a new 4200 with 24.03-RELEASE and noticed that my PORT2LAN (igc2) on the device is flapping. I noticed that the port was only negotiating at 100baseTX <full-duplex> instead of 1000baseTX <full-duplex> as well. Other observations I noticed is that the link speed light on the right of the port is off and only the left light is blinking green.
PORT2LAN is plugged into a Netgear 16 port unmanaged switch. The speed light on the switch shows as 10/100 link speed as well.
Looking at the logs, you see a lot of this flapping in a short period of time:
igc2: link state changed to UP igc2: link state changed to DOWN igc2: link state changed to UP igc2: link state changed to DOWN igc2: link state changed to UP igc2: link state changed to DOWN igc2: link state changed to UP igc2: link state changed to DOWN igc2: link state changed to UP igc2: link state changed to DOWN igc2: link state changed to UP igc2: link state changed to DOWN igc2: link state changed to UP
More detailed logs showing the flapping:
Aug 18 01:48:19 pfSense php-fpm[23304]: /rc.linkup: Hotplug event detected for PORT2LAN(lan) dynamic IP address (4: 192.168.1.1, 6: track6) Aug 18 01:48:19 pfSense php-fpm[23304]: /rc.linkup: DEVD Ethernet detached event for lan Aug 18 01:48:23 pfSense check_reload_status[618]: Reloading filter Aug 18 01:48:23 pfSense check_reload_status[618]: Reloading filter Aug 18 01:48:27 pfSense check_reload_status[618]: Linkup starting igc2 Aug 18 01:48:27 pfSense kernel: igc2: link state changed to UP Aug 18 01:48:28 pfSense php-fpm[95339]: /rc.linkup: Hotplug event detected for PORT2LAN(lan) dynamic IP address (4: 192.168.1.1, 6: track6) Aug 18 01:48:28 pfSense php-fpm[95339]: /rc.linkup: DEVD Ethernet attached event for lan Aug 18 01:48:28 pfSense php-fpm[95339]: /rc.linkup: HOTPLUG: Configuring interface lan Aug 18 01:48:28 pfSense check_reload_status[618]: Linkup starting igc2 Aug 18 01:48:28 pfSense php-fpm[95339]: /rc.linkup: Gateway, NONE AVAILABLE Aug 18 01:48:28 pfSense php-fpm[95339]: /rc.linkup: Gateway, NONE AVAILABLE Aug 18 01:48:28 pfSense php-fpm[554]: /rc.newwanipv6: rc.newwanipv6: Info: starting on igc3. Aug 18 01:48:28 pfSense php-fpm[554]: /rc.newwanipv6: rc.newwanipv6: No IPv6 address found for interface PORT1WAN [wan]. Aug 18 01:48:28 pfSense check_reload_status[618]: Restarting IPsec tunnels Aug 18 01:48:28 pfSense php-fpm[554]: /rc.newwanipv6: rc.newwanipv6: Info: starting on igc3. Aug 18 01:48:28 pfSense php-fpm[554]: /rc.newwanipv6: rc.newwanipv6: No IPv6 address found for interface PORT1WAN [wan]. Aug 18 01:48:28 pfSense kernel: igc2: link state changed to DOWN Aug 18 01:48:30 pfSense php-fpm[5106]: /rc.newwanipv6: rc.newwanipv6: Info: starting on igc3. Aug 18 01:48:30 pfSense php-fpm[5106]: /rc.newwanipv6: rc.newwanipv6: on (IP address: 2600:8804:3f00:600:49a0:7a16:f6e3:e8f1) (interface: wan) (real interface: igc3). Aug 18 01:48:31 pfSense php-fpm[95339]: /rc.linkup: The command '/usr/local/sbin/unbound -c /var/unbound/unbound.conf' returned exit code '1', the output was '[1723963711] unbound[26875:0] error: bind: address already in use [1723963711] unbound[26875:0] fatal error: could not open ports' Aug 18 01:48:32 pfSense php-fpm[5106]: /rc.newwanipv6: Gateway, NONE AVAILABLE Aug 18 01:48:32 pfSense php-fpm[5106]: /rc.newwanipv6: Gateway, NONE AVAILABLE Aug 18 01:48:32 pfSense check_reload_status[618]: Reloading filter Aug 18 01:48:34 pfSense check_reload_status[618]: updating dyndns lan Aug 18 01:48:34 pfSense check_reload_status[618]: Reloading filter Aug 18 01:48:34 pfSense check_reload_status[618]: Reloading filter Aug 18 01:48:34 pfSense php-fpm[554]: /rc.linkup: Hotplug event detected for PORT2LAN(lan) dynamic IP address (4: 192.168.1.1, 6: track6) Aug 18 01:48:34 pfSense php-fpm[554]: /rc.linkup: DEVD Ethernet detached event for lan Aug 18 01:48:37 pfSense check_reload_status[618]: Reloading filter Aug 18 01:48:37 pfSense check_reload_status[618]: Reloading filter Aug 18 01:48:41 pfSense check_reload_status[618]: Linkup starting igc2 Aug 18 01:48:41 pfSense kernel: igc2: link state changed to UP
I have tried using a new cable but experienced the same results. I did reboot the router after seeing this behavior and the flapping stopped, but the port is still negotiating at 100baseTX <full-duplex>.
I did create a support ticket and the rep suggested I try one of the OPT ports to see if this behavior exists there. I'm going to try that, but wanted to get this post out created to see if anyone has experienced anything similar.
-
UPDATE
I was able to configure PORT3 as the new LAN port on my router. It's negotiating correctly at 1000baseT <full-duplex>.
However, PORT2LAN is still negotiating at 100baseTX <full-duplex>. Still not sure why and how this can be fixed.
-
@i298X9wY1jGQ Running at 100M is typically a sign that one wire-pair within the ethernet cable is not connecting. If you already replaced the cable (and are sure the replacement is good), the next most likely theory is a dirty connector pin within the port. Try cleaning the pins. If that doesn't help, it's a faulty port.
-
@tgl said in 4200 - LAN port flapping:
@i298X9wY1jGQ Running at 100M is typically a sign that one wire-pair within the ethernet cable is not connecting. If you already replaced the cable (and are sure the replacement is good), the next most likely theory is a dirty connector pin within the port. Try cleaning the pins. If that doesn't help, it's a faulty port.
Yeah, tried cleaning the pins even though they looked very clean since this is a brand new unit. However, after rebooting the unit, the port stays active but only at 100M. I'm not sure if this is a faulty port or something funky going on with the software.
-
Do you have the port set to a specific speed? On the 4200 it should pretty much always be set to 'default'.
-
@stephenw10 said in 4200 - LAN port flapping:
Do you have the port set to a specific speed? On the 4200 it should pretty much always be set to 'default'.
I tried forcing my laptop to use 1G full-duplex but PORT 2 still auto negotiates at 100M. Then, I tried forcing PORT 2 to negotiate at 1G and it is now showing "no carrier" and is completely off (no activity lights). After making the change to negotiate at a specific speed always makes the port unresponsive and requires a reboot for it to become active again.
I'm on my third cable now and can probably rule out any issues with the cables I'm using.
-
Looking at the 4200 documentation, there's no section for the RJ-45 LED activity lights so I'm not sure how to interpret the colors. I see other models have documentation going over the activity lights. Examples are here and here.
-
@i298X9wY1jGQ said in 4200 - LAN port flapping:
Looking at the 4200 documentation, there's no section for the RJ-45 LED activity lights so I'm not sure how to interpret the colors.
Yeah, that seems like a documentation oversight. However, "ifconfig" in the shell will tell you the truth about port speeds. On my 4200 I see
media: Ethernet autoselect (2500Base-T <full-duplex>) ... media: Ethernet autoselect (1000baseT <full-duplex>)
on my two active ports, and those are both right.
-
@tgl said in 4200 - LAN port flapping:
@i298X9wY1jGQ said in 4200 - LAN port flapping:
Looking at the 4200 documentation, there's no section for the RJ-45 LED activity lights so I'm not sure how to interpret the colors.
Yeah, that seems like a documentation oversight. However, "ifconfig" in the shell will tell you the truth about port speeds. On my 4200 I see
media: Ethernet autoselect (2500Base-T <full-duplex>) ... media: Ethernet autoselect (1000baseT <full-duplex>)
on my two active ports, and those are both right.
I ended up reinstalling pfSense fresh. Confirmed in the shell that PORT2 was still showing as 100baseTX <full-duplex>. Not really sure what else to do here.
-
@i298X9wY1jGQ said in 4200 - LAN port flapping:
Not really sure what else to do here.
If this is a new unit, it sounds like RMA time. I think you've covered every possibility other than faulty port hardware.
-
@tgl said in 4200 - LAN port flapping:
@i298X9wY1jGQ said in 4200 - LAN port flapping:
Not really sure what else to do here.
If this is a new unit, it sounds like RMA time. I think you've covered every possibility other than faulty port hardware.
Yeah, the rep I had been working with initiated an RMA. I was really hoping I would not have to do that but it is what it is.
Thanks all for the help!
-
As a test try just linking two ports together directly and then check the link state at the CLI with
ifconfig -v
.There shouldn't be an issue linking at 1G. You may have a bad port.