Port constantly goes up/down

hi, i have problem with some ports goes up and down:

[200765.135528] Atheros AR8216/AR8236/AR8316 mdio.0:00: Port 2 is down
[238863.330478] Atheros AR8216/AR8236/AR8316 mdio.0:00: Port 5 is down
[238866.450024] Atheros AR8216/AR8236/AR8316 mdio.0:00: Port 5 is up
[240705.164462] Atheros AR8216/AR8236/AR8316 mdio.0:00: Port 5 is down
[240706.204512] Atheros AR8216/AR8236/AR8316 mdio.0:00: Port 5 is up
[243022.277137] Atheros AR8216/AR8236/AR8316 mdio.0:00: Port 5 is down
[243024.356752] Atheros AR8216/AR8236/AR8316 mdio.0:00: Port 5 is up
[247297.703155] Atheros AR8216/AR8236/AR8316 mdio.0:00: Port 4 is down
[247299.782717] Atheros AR8216/AR8236/AR8316 mdio.0:00: Port 4 is up
[247961.220260] Atheros AR8216/AR8236/AR8316 mdio.0:00: Port 2 is up

and before it starts there is some strange dmesg messages:

[18860.379450] ath10k_pci 0000:00:00.0: htt tx: fixing invalid VHT TX rate code 0xff
[24773.758367] ath10k_pci 0000:00:00.0: wmi: fixing invalid VHT TX rate code 0xff

i having same problem on stock firmware, and sometimes for some reason internet is randomly in time just stop working

tested on all devices android/windows/fedora on all port up/down so it problem with router

Device:
Tp-Link Archer C6 V2 EU
OpenWrt Version: 22.03.2 r19803-9a599fee93
Luci: openwrt-22.03 branch git-22.288.45147-96ec0cd
Kernel: 5.10.146

The port up/down messages are related to the switch, the ath10k_pci messages are related to WiFi.

If only a specific port goes up/down, I would check/replace the cable first. However, since you are experiencing the same behavior on stock, it is also possible that your device is faulty.

I would think about a faulty patch cord or RJ45 first, but if several ports are going up and down often, are you sure you're using Cat6 or at least Cat5E cables? Else, maybe the power brick is faulty, or the router itself. If you're not sure, one test you could run would it be to force the port(s) to 100Mbps for a while.

@andyboeh its not cable problem i mentioned that its problem on all ports it happens randomly, and with wifi i have strange behavior, sometimes 2.4ghz wifi just stops working. devices can't see it and 5ghz visible on wifi list but can't connect or it instantly disconnect you or it just freezes on 'authehicating', and problem with ports goes up/down somehow releated to ath10k_pci messages, and sometimes internet speed so slow(1mbs download 90mbs upload) (speed of provider 100mbps)

@feliciano i using cat5e cables, i tried iperf3 to router(it gives 300mb/s) and when some device using 60+ mbps of internet other devices speed is 10mb/s, issue with port goes up/down happends randomly, independent of internet usage

and sometimes for no reason ram is just full and cpu can't handle but devices not using internet at all

and sometimes internet speed is very slowly drops(10mb/s per hour) until it will be so low and can't be used

oh and sometimes after reboot there is no internet even speed of iperf3 to router is 5mbps/s and cpu usage is 100% and i get spammed in console by this messages:

# this messages from logread
br-lan: received packet on eth0.1 with own address as source address (addr:60:32:b1:ff:55:bd, vlan:0)
DHCPDECLINE(br-lan) 192.168.1.215 00:02:2a:e1:7f:6e # (its not only on 192.168.1.215 it was spammed with different ip addreses)

# oh and i get spammed with:
DHCPDISCOVER(br-lan) 00:02:2a:e1:7f:6e
DHCPOFFER(br-lan) 192.168.1.113 00:02:2a:e1:7f:6e
DHCPDECLINE(br-lan) 192.168.1.133 00:02:2a:e1:7f:6e # this messages is normal but normarly i don't get 'DHCPDECLINE' and its spammed 5 messages/1s

# dmesg: i receive same messages as in logread but also i receive this:
net_ratelimit: 105355 callbacks suppressed

What devices are connected to the Ethernet ports?

Windows 7 PC, Windows 11 PC, Fedora Linux 37 PC(Why you need this information?)

Because I am curious if any of those devices could actually be responsible for the issue. Do any of those systems use a USB-C docking hub with ethernet?

No, this devices directly connected with motherboard port, without any hubs/or other things

Ok... with that, I'd say that it is unlikely that these hosts are causing the problem.... just wanted to rule that possibility in/out.

This makes it sound like the hardware itself may have an issue. A faulty/failing power supply could actually produce these types of issues if it browns-out the main power system in the router. Or it could be internal with bad capacitors, if it is indeed a hardware issue.

But we could take a look at your configuration to make sure nothing there is likely to cause issues.

Windows will shut down the PC's Ethernet port for a few seconds while changing the speed (10 Mb uses less power than Gb) every time it goes to sleep or wakes up. This will create the up/down log entries at the switch which are completely expected and normal.

The device with the 00:02 MAC address in the logs you posted is misbehaving in a few ways. Identify it and fix or remove from the network.

1 Like