[2.1.5-RELEASE] Bouncy PPPoE WAN (BT Home Broadband/Bridged TP-LINK TD-8817)

stephenw10

Hmm, well apinger is still monitoring 8.8.8.8. but it goes straight to 'down' rather than delay or packet loss alarms. Do you still see ping times and packet loss stats in Status: Gateways: ?

Steve

firewalluser

Try an EU based ip address?

I have a Swiss ip address in the Monitor IP field for ppoe and have not had a problem with net connections dropping on TalkTalk in the UK, although for me, I was on 2.1.4 when this started happening a short while back. I moved to 2.2 beta but still saw the same problem, but only after putting an ip address in the Monitor IP field did my net connections stay up and come back up when the connection was eventually reset and new ip address assigned. Doing the same might work for you on 2.1.4.

This only became a problem a month or so ago so I suspect some changes have occurred with UK net infrastructure recently.

phil.davis

Dec 11 21:12:04	ppp: [wan_link0] LCP: no reply to 1 echo request(s)
Dec 11 21:12:14	ppp: [wan_link0] LCP: no reply to 2 echo request(s)
Dec 11 21:12:24	ppp: [wan_link0] LCP: no reply to 3 echo request(s)
Dec 11 21:12:34	ppp: [wan_link0] LCP: no reply to 4 echo request(s)
Dec 11 21:12:44	ppp: [wan_link0] LCP: no reply to 5 echo request(s)
Dec 11 21:12:44	ppp: [wan_link0] LCP: peer not responding to echo requests
Dec 11 21:12:44	ppp: [wan_link0] LCP: state change Opened --> Stopping

I will just point out that Link Control Protocol (LCP) are control packets embedded in PPP to control and monitor the link just from your device (pfSense in this case) to the ISP's connection point - just a single point-to-point hop.
http://en.wikipedia.org/wiki/Link_Control_Protocol

These are like "ping" at a lower and more local level. It is saying that the ISP end has stopped responding.
Since putting Windows or some other device in place of pfSense works, then I guess there (must?) be some interoperability thing between the implementation of PPP+LCP FreeBSD/pfSense and in Windows. It would require some detailed capture of exactly what is going across the wire in PPP packets to analyse that.

Having given no practical pointer to a solution, I shall press "Post" and go to sleep :(

stephenw10

:) Good point.
I don't think we've had a definitive answer as to whether it stays up using the Windows PPPoE client yet.
What we have had is a number of posts on the forum complaining about bad BT connections in recent days. Coincidence? ;)

Steve

joe91

@stephenw10:

:) Good point.
I don't think we've had a definitive answer as to whether it stays up using the Windows PPPoE client yet.
What we have had is a number of posts on the forum complaining about bad BT connections in recent days. Coincidence? ;)

Steve

Sorry for the delay in responding, I've been trying different things, currently back to HomeHub with a stable connection, as I have tried rebuilding the PPPoE connection, fresh install of pfSense, swapping the NICs round all still result in bouncy connections.

All tests I have done have been stable when connected directly to Windows PC, but I have not had chance to leave this for a whole day yet. I will be away for a few weeks and need reliable access back into the network, so will start trying to resolve this again in the new year.

Any further ideas welcome.

firewalluser

You dont say which BT router you are using.

I know the BT Business 2wire router has bugs in the software which forced me to abandon it when you run it in bridge mode,, namely if it lost power your settings would be lost plus BT can also remotely override your settings, so if you switched off the wifi on the router, you would see this being switched back on usually within a day but could be as quickly as a few hours.

http://www.nairod.co.uk/mountain-lion-download-problems-and-bt-business-2-wire-routers/

joe91

@firewalluser:

You dont say which BT router you are using.

I know the BT Business 2wire router has bugs in the software which forced me to abandon it when you run it in bridge mode,, namely if it lost power your settings would be lost plus BT can also remotely override your settings, so if you switched off the wifi on the router, you would see this being switched back on usually within a day but could be as quickly as a few hours.

http://www.nairod.co.uk/mountain-lion-download-problems-and-bt-business-2-wire-routers/

Sorry thought I'd mentioned earlier or in OP. It's a HomeHub4 it has no concept of bridge mode or I'd just use that

stephenw10

As Phil pointed out earlier you are seeing a failure of the PPPoE connection at a low level. I would be suspicious of the tp-link device which has to bridge the pppoe to pppoa. I think you will have to prove that device by connecting through it with something else. Either directly through Windos pppoe client or the homehib can do pppoe if want to try that.

Steve

joe91

@stephenw10:

As Phil pointed out earlier you are seeing a failure of the PPPoE connection at a low level. I would be suspicious of the tp-link device which has to bridge the pppoe to pppoa. I think you will have to prove that device by connecting through it with something else. Either directly through Windos pppoe client or the homehib can do pppoe if want to try that.

Steve

Hi everyone, sorry for the long delay in getting back to you, been very busy, so I've moved to Plusnet to get a Static IP and half my bill, still experiencing same problems. I have upgraded to 2.2-RELEASE (amd64).

Earlier I connected directly using the TP-Link and plugged the Ethernet into my Asus RT-AC68U. Within 30 minutes the same problem occurred and did not reconnect after 10 minutes, now back in bridge mode and plugged back into pfSense.

The log from the TP-Link says


2/25/2015 12:19:20> Last errorlog repeat 6 Times
2/25/2015 12:19:20> Call Failed
2/25/2015 12:19:20> netMakeChannDial: err=-3000 rn_p=8043336c
2/25/2015 12:19:26> Last errorlog repeat 15 Times
2/25/2015 12:19:27> netMakeChannDial: err=-3000 rn_p=8043336c
2/25/2015 12:19:27> Last errorlog repeat 1 Times
2/25/2015 12:19:28> netMakeChannDial: err=-3000 rn_p=8043336c
2/25/2015 12:19:38> Last errorlog repeat 16 Times
2/25/2015 12:19:39> netMakeChannDial: err=-3000 rn_p=8043336c
2/25/2015 12:19:39> Last errorlog repeat 2 Times
2/25/2015 12:19:41> netMakeChannDial: err=-3000 rn_p=8043336c
2/25/2015 12:19:43> Last errorlog repeat 3 Times
2/25/2015 12:19:44> netMakeChannDial: err=-3000 rn_p=8043336c
2/25/2015 12:19:51> Last errorlog repeat 20 Times
2/25/2015 12:19:51> Call Failed
2/25/2015 12:19:52> netMakeChannDial: err=-3000 rn_p=8043336c

Must have been luck when worked on Windows, or Windows is better at reconnecting quickly.

Any recommendations? I can think of two:
RMA TP-Link for replacement
RMA TP-Link for Return and get a different Modem (any suggestions on which?)

stephenw10

Hmm, interesting. At least you've proved it's not a pfSense specific issue.
I'm on Plusnet and seeing very few issues. However if you are you should get on to their support forum and ask because they have real people with real technical knowledge there who will help. One of their support guys went to amazing lengths to pin down a fault I found.

Are you on fibre? I used a V120 with their adsl2 service for years.

Steve

joe91

@stephenw10:

Hmm, interesting. At least you've proved it's not a pfSense specific issue.
I'm on Plusnet and seeing very few issues. However if you are you should get on to their support forum and ask because they have real people with real technical knowledge there who will help. One of their support guys went to amazing lengths to pin down a fault I found.

Are you on fibre? I used a V120 with their adsl2 service for years.

Steve

Thanks Steve, moving discussion over to Plusnet Support Forum http://community.plus.net/forum/index.php/topic,137223.0.html