FS#3280 - Wan interface and sometimes lan interfaces randomly go up/down at the link layer

[openwrt-bot]

Dynamo:

• Versions where bug is present: OpenWRT snapshots, tested with r13938 and r18039.

• Affected device: D-Link DIR-878 and likely others from its family.

• What does it do and shouldn't do: the ethernet interfaces randomly get switched on/off, dropping the link and resetting the connection. Troubleshooting has already been thoroughly conducted to rule out other causes, see below. This is critical severity because makes OpenWRT unusable in these devices.

• Steps to reproduce:
  -- flash the snapshot image on a D-Link DIR-878;
  -- configure the WAN (tested with PPPoE and DHCP protocols) and LAN interfaces (tested with Static and DHCP protocols) - using all interfaces will increase the likelihood that at least one of them will present the behaviour;
  -- wait an indefinite but reasonable (< 2h for all instances so far) amount of time;
  -- the interface will randomly be put down and then up a couple seconds later, and it will be randomly looping up/down states at random amounts of time.

• Troubleshooting already conducted: full and detailed troubleshooting history is present at [[https://forum.openwrt.org/t/pppoe-disconnects-all-the-time/71165|this]] forum thread. Please let me know if I can provide further information.

• Additional info: [[https://forum.openwrt.org/t/pppoe-disconnects-all-the-time/71165/18|some]] [[https://forum.openwrt.org/t/dir-882-possible-wi-fi-issues-with-new-mt76-drivers/71292|people]] are having WiFi issues with MT7621 family devices that might be related to this one if they share a common cause, such as a possible race condition in the drivers;

[openwrt-bot]

Dynamo:

The Archer C60 device appears to also be affected:
https://bugs.openwrt.org/index.php?do=details&task_id=3271
https://forum.openwrt.org/t/archer-c60-wan-disconnects/68896
https://forum.openwrt.org/t/udhcpc-does-not-work-correctly/69774

[openwrt-bot]

francocastillo:

Yes, that's why I had to go back to 19.07.3 that this problem does not happen.

[openwrt-bot]

alpianon:

I had exactly the same problem with 19.07.7 on a TP-Link Archer C7 v5. I reverted to 19.07.3 and the problem is gone, connection works perfectly

[openwrt-bot]

ev6ds:

I have the same issue with a WRC-2533GST2 both with the snapshot firmware and the newer 21.02.0 firmware from https://openwrt.org/toh/hwdata/elecom/elecom_wrc-2533gst2
Is there anyway to tell what version of the firmware was included in the snapshot from that page to know if reverting to 19.07.3 would help me?
19.07.3 looks like it was released Wed May 20 04:10:19 2020
And the snapshot looks like it was added 2020/06/27 09:05.
Do we think the cause of the issue was introduced during that month?
Maybe I can backport the patch and rebuild a version of 19.07.3 for my device if it fixes it.

[openwrt-bot]

Dynamo:

This is fixed for me on 21.02.0-rc4.

[openwrt-bot]

ev6ds:

Yeah I took a quick look at the logs and it seems it did indeed get fixed with 21.02.0 for me. The new error I got was slightly different and likely the result of a provider issue that resolved itself. So glad this is fixed, thanks guys!