Turris Omnia drops WAN occasionally

I’m running Turris Omnia 2019 with TurrisOS 4.0.5 in PPPoE WAN mode behind a DrayTek Vigor 130 DSL modem. Everything works fine except one thing: the WAN connection just drops sometimes and can’t be neither reestablished automatically nor due to single restarts of both devices.

It drops randomly. It has happend around one week after I’ve taken the setup into production for the first time. Then there was another drop one week later and just another one a day later. No special times detected, clock-wise.

When I read the syslog via logread I see these messages being displayed in a endless loop:

Feb  9 15:31:49 turris netifd: Interface 'wan' is enabled
Feb  9 15:31:49 turris netifd: Interface 'wan' is setting up now
Feb  9 15:31:49 turris insmod: module is already loaded - slhc
Feb  9 15:31:49 turris insmod: module is already loaded - ppp_generic
Feb  9 15:31:49 turris insmod: module is already loaded - pppox
Feb  9 15:31:49 turris insmod: module is already loaded - pppoe
Feb  9 15:31:49 turris pppd[20604]: Plugin rp-pppoe.so loaded.
Feb  9 15:31:49 turris pppd[20604]: RP-PPPoE plugin version 3.8p compiled against pppd 2.4.7
Feb  9 15:31:49 turris pppd[20604]: pppd 2.4.7 started by root, uid 0
Feb  9 15:31:49 turris kresd[20608]: [priming] cannot resolve '.' NS, next priming query in 10 seconds
Feb  9 15:31:49 turris kresd[20608]: [detect_time_skew] cannot resolve '.' NS
Feb  9 16:31:50 turris kernel: [ 1654.714951] mvneta f1034000.ethernet eth2: Link is Down
Feb  9 15:31:50 turris netifd: Network device 'eth2' link is down
Feb  9 15:31:50 turris netifd: Interface 'wan' has link connectivity loss
Feb  9 15:31:52 turris netifd: Network device 'eth2' link is up
Feb  9 15:31:52 turris netifd: Interface 'wan' has link connectivity 
Feb  9 15:31:52 turris netifd: Interface 'wan' is setting up now
Feb  9 16:31:52 turris kernel: [ 1656.791736] mvneta f1034000.ethernet eth2: Link is Up - 1Gbps/Full - flow control rx/tx

It seems that restarting the modem solves the issue most of the time, but the modem itself doesn’t seem to have a problem, because when I connect to it directly, it says it has a valid DSL connection. I randomly restart the devices and after some tries it works again.

had more or less the same issue - turned off the media detection (?) for that connection, job done.

Can you tell me where I can do that?

Probably force_link [1]? But that would not be recommended for PPPoE

Specifies whether ip address, route, and optionally gateway are assigned to the interface regardless of the link being active (‘1’) or only after the link has become active (‘0’); when set to ‘1’, carrier sense events do not invoke hotplug handlers

IMHO netifd suffers issues with media sensing which may not exhibit with more contemporary versions.


[1] https://openwrt.org/docs/guide-user/base-system/basic-networking

Agreed. It’s a bit rude. But didn’t find anything more sensible. And - it helped.

This with PPPoE on the subscriber line?

Yes. Different modem as the thread opener, a netgear.

I will try to do that the next time the problem occurs and tell you about it afterwards. Thank you so far!

And it has occured just today again. Checked the Force link option and rebooted the device. Connection was reestablished directly after that. I will watch the behaviour for the next two weeks and will report if this issue has been solved by checking this option or not! Thank you again!