Archer C7v4 random eth0: link down

I currently face severe trouble with an Archer C7v4 and OpenWrt 18.06.1 (same on 18.06.0). The router itselfes is brand new. After a couple of hours / days (it is really quite random) it stops working properly. Wired clients can't connect anymore, via wireless clients I am still able to connect to LuCi and access the logs.

Logs do not seem to contain anything weird (despite the error). The router ends up with eth0: link down.

System log:

Mon Sep  3 07:42:55 2018 kern.info kernel: [119774.409536] Atheros AR8216/AR8236/AR8316 ag71xx-mdio.0:00: Port 1 is up
Mon Sep  3 07:42:55 2018 kern.info kernel: [119774.416989] Atheros AR8216/AR8236/AR8316 ag71xx-mdio.0:00: Port 2 is up
Mon Sep  3 07:42:55 2018 kern.info kernel: [119774.424448] Atheros AR8216/AR8236/AR8316 ag71xx-mdio.0:00: Port 3 is up
Mon Sep  3 07:42:55 2018 kern.info kernel: [119774.431941] Atheros AR8216/AR8236/AR8316 ag71xx-mdio.0:00: Port 4 is up
Mon Sep  3 07:42:55 2018 kern.info kernel: [119774.440558] eth0: link down
Mon Sep  3 07:42:55 2018 daemon.notice netifd: Network device 'eth0' link is down
Mon Sep  3 07:42:55 2018 daemon.notice netifd: VLAN 'eth0.1' link is down
Mon Sep  3 07:42:55 2018 daemon.notice netifd: VLAN 'eth0.2' link is down
Mon Sep  3 07:42:55 2018 daemon.notice netifd: Interface 'wan' has link connectivity loss
Mon Sep  3 07:42:55 2018 daemon.notice netifd: Interface 'wan6' has link connectivity loss
Mon Sep  3 07:42:55 2018 kern.info kernel: [119774.450820] br-lan: port 1(eth0.1) entered disabled state

Kernel log:

[119774.409536] Atheros AR8216/AR8236/AR8316 ag71xx-mdio.0:00: Port 1 is up
[119774.416989] Atheros AR8216/AR8236/AR8316 ag71xx-mdio.0:00: Port 2 is up
[119774.424448] Atheros AR8216/AR8236/AR8316 ag71xx-mdio.0:00: Port 3 is up
[119774.431941] Atheros AR8216/AR8236/AR8316 ag71xx-mdio.0:00: Port 4 is up
[119774.440558] eth0: link down
[119774.450820] br-lan: port 1(eth0.1) entered disabled state

Up to now I tried tons of things, but did not find a solution. Any ideas?

Perhaps related to

1 Like

Unfortunately had to switch routers. On the Archer C7v4 I get no more than 1 day uptime, then "eth0: link down" (sometimes no more than half a day). Happy to test with newer OpenWrt versions, but for now the v4 seems to be unusable in my case.

Sorry it didn't work with the C7v4, what hardware/firmware did you switch to and has it been stable?

After using various TP-Links during the last couple of years (was really happy with all of them, really a pitty that C7v4 does not seem to be reliable with current builds) I switched back in the Linksys camp and now use a WRT3200ACM. For now (up since ~1 week) it works great in my environment.

Thanks for the update on this wjwj, glad you have a working solution.

I was just about to post a question to you about how your wired network is set up, as the base problem described in this patch: https://git.openwrt.org/?p=openwrt/openwrt.git;a=commit;h=f4f99ec9737c653815268f2efad0210caaa32e2d was that link state changes or speed changes could trigger a hardware bug in the SoC to physical device link and cause of loss of LAN access until that link is reset.

If you had multiple devices physically attached to the LAN ports, one or more of them (or their interaction) could have been the cause. Isolating them with a separate switch might have prevented the bug from manifesting.

If you still have the C7v4, a quick test of that would be helpful.

Good to know you are working.

A big +1 to the request TopDog posted, sounds like an interesting workaround until someone takes that linked patch and adapts it for the C7.
If you were able to prove the external switch prevented the LAN drops, it would be a great indicator to include in a new bug report and see if we can get this fixed.

@TopDog: This was also my best guess, but since figuring out how to build OpenWrt with this patch enabled for the C7v4 is out of scope for me due to time constraints I did not test.

I had devices connected to all ports on the router. What I recognized (also currently on the WRT3200ACM) is my AppleTV randomly switching quite randomly between 1000 and 100 MBit.

I‘ll try to isolate the device and test. Also happy to test if you have a build ready with the patch included.

@wjwj Thanks for the additional detail, especially re: the Apple TV, guessing it flips into 100Mbps mode when asleep and back up to Gig when 'awake' (even if not 'in use') to check with the cloud services.
Laptops and desktop PCs also do similar things, like switching down to 10Mbps for sleep with WOL enabled.

Dev's, anyone able to create a patch for the C7v4 (and or v2) and I'll be happy to test as well.

I have a spare C7 I can run cascaded to my main router and plug in a variety of devices, like the Apple TV.

1 Like

Hello,

I have new TP-LINK Archer C7 v5 and get exactly the same behavior e.g. loss of connectivity on all ethernet ports (WAN ad LAN), while WiFi still work. I can access the router via wi-fi but can't connect to computers on LAN nor to Internet.

Using lates firmware.

Model: TP-LINK Archer C7 v5
Architecture: Qualcomm Atheros QCA956X ver 1 rev 0
Firmware Version: OpenWrt 18.06.2 r7676-cddd7b4c77 / LuCI openwrt-18.06 branch (git-19.020.41695-6f6641d)
Kernel Version: 4.9.152

Kernel log gives following:
[277429.566846] Atheros AR8216/AR8236/AR8316 ag71xx-mdio.0:00: Port 1 is up
[277429.574295] Atheros AR8216/AR8236/AR8316 ag71xx-mdio.0:00: Port 2 is up
[277429.582270] Atheros AR8216/AR8236/AR8316 ag71xx-mdio.0:00: Port 4 is up
[277429.590788] eth0: link down
[277429.595459] br-lan: port 1(eth0.1) entered disabled state

Is there any progress on this issue?

Aleš

Unfortunately, until a dev integrates the patch I listed in this prior post ( Archer C7v4 random eth0: link down ) I don't think we'll see an improvement.

Anyone have suggestions for how to escalate this? Or even for what to try to DIY?

I have same issue.

Model: WZR-HP-AG300H
Architecture: Atheros AR7161 rev 2
Version: OpenWrt 18.06.1, 18.06.4, 18.06.5

Kernel log:

[   54.587197] random: crng init done
[543767.035100] Atheros AR8216/AR8236/AR8316 ag71xx-mdio.0:04: Port 4 is up
[543767.041977] eth1: link down
[543769.115216] ar71xx: pll_reg 0xb8050014: 0x110000
[543769.119922] eth1: link up (1000Mbps/Full duplex)