DSL Modem disconnects and the RT-N66U

[Gordon1M]

CONFIGURATION

• ISP: Bell Canada DSL Business Internet (Fibe Internet 10)
• Hub: Bell Home Hub 1000 (Sagemcom F@st 4350). The problems originally started while using an old Speedstream modem, which was replaced by Bell Canada with the HH1000.
• Second router: Asus RT-N66U. Firmware is Merlin:378.54_2.
• The HH1000 is operating in Normal (not Bridge) mode, so it uses both the modem and router functions of the Hub. The WAN port of the N66U is connected to one of the HH1000 LAN ports. The HH1000 DHCP server is running and the N66U is set to Autoconnect (not PPPoE).

PROBLEM

I have a problem with apparently random and intermittent Internet disconnects, resulting in pauses of a few seconds or outages of several hours. This the Hub/modem disconnecting, not the router. I have not noticed any occasions when the router has become disconnected from the HH1000. The modem connection may be stable for a few days or there may be several disconnects an hour. These disconnects started in April this year. Typically, during disconnects, there are multiple PPPoE timeout errors. I have seen several hours of these error messages in the HH1000 System Log; once every couple of seconds. Before the problem started in April, the connection had been running for about a year with no apparent problems.

A visiting Bell tech claimed that Asus routers don't work well with the HH1000. He didn't solve the problem.

I noted the advice about using Cat5e or Cat 6 cable in another recent post, but I think that this likely would only affect router to Hub (router) connections and not Hub (modem) to ISP connections.

I don't see how the N66U can cause the modem to disconnect in this configuration. Am I correct?

Thank you
Gordon

 

[RMerlin]

To be blunt, Bell's Sagemcom modem/routers are garbage. Even Bell themselves knows it - one of my customers who had stability issues was told by Bell that they should bridge the Sagemcom and use their own router. (Except that this tech support employe obviously didn't know that Bell removes the ability to turn their Sagemcom into a bridge...)

I know that Bell pushed firmware upgrades a few months ago which were supposed to at least resolve the old "Sync and no surf" issues. Dunno if the general stability was also improved then.

I think the Sagemcom has a mode called "Super DMZ" or something like that (I only worked once or twice with these models). In that mode, you can have it work almost as a bridge, and have your router handle the PPPoE authentication, which should be much more stable.

Have a look at the Bell forums at DLSReports, you might be able to get more tips there from other Bell users.

 

[Gordon1M]

Thanks Merlin. I have tried the HH1000 in bridge mode (this can be done on request by Bell) and there was no difference in the disconnections. Of course, this immediately results in Bell saying that the problem lies with the N66U, which then has to make the PPPoE connection.

I'll check for any firmware upgrades for the HH1000 and also the "Super DMZ".

I have looked at the Bell DSLReports forums and have, in fact, had some conversations with a Bell person on the "private" BellDirect forum. This has been inconclusive so far, though I am awaiting another response.

The network location in question is local , though remote from my home, so I can't immediately investigate disconnections. I have arranged for the N66U to e-mail the WAN IP address to me every hour and whenever the WAN comes up. This at least tells me when the connection has gone down. It went down between 3 and 4.00pm today and is still down. I went into the location, when this happened, and the WAN is showing as down in the modem stats and by the red light on the modem. However, there is nothing negative showing up in the modem system log (which has previously shown PPPoE time outs. I managed to get the Bell tech who visited today to check the line after the connection was lost and was told that the line was OK .... I suspect that the line checks are fairly basic.

Anyway, you don't seem to be suggesting that I should look at the N66U for the problem, which was the main point of my post.

Thank you
Gordon

 

[RMerlin]

Anyway, you don't seem to be suggesting that I should look at the N66U for the problem, which was the main point of my post.

I wouldn't rule out a router-specific issue, but I doubt it's a general issue rather than a configuration or hardware issue. I have quite a few customers with Bell who have various Asus routers, from the RT-N12 up to an RT-AC56U, and none of them have any stability issue.

It could also be a modem issue, though I just saw that the issue started back to the original Speedstream modem. Next thing I would suggest then is swapping phone wire (even move to a different plug if possible) and the Ethernet cable between the modem and the router, if you haven't done so yet.

DSL issues are always a PITA to diagnose, especially those that are random. Had a few customers who had random issues caused by the old copper wiring outside, which would become unreliable specifically during the spring (as melting water entered the poorly insulated wires) or very windy days causing wires to jiggle too much. Those can unfortunately not be observed by a line test, as it's not a constant SNR problem. If you can get access to another compatible modem (maybe from a friend), it might be worth swapping modems for a few days to see if the problem moves to the other person's connection.

 

[RMerlin]

PS: two of those customers of mine who had random issues that Bell never managed to address (even tho it was obvious the outside wiring was at fault - it was even visible), we ended up moving both of them to cable. Zero issues so far, after close to two years for each of them.

 

[Gordon1M]

Thanks again Merlin. I have tried to eliminate the various components. In fact, some time ago, I temporarily changed the router to a WRT54GL and disconnects still occurred, though not to the extent that I am seeing now (As of this morning, the WAN connection is down and has been in this state since 3pm yesterday). I suspect that I am going to have to power cycle the hub to obtain a connection. This situation is much worse than it used to be. This seems to suggest a problem with the line or the hub to be the most likely. I have resisted blaming the hub for the problem, given that I had originally experienced the problem (though less extreme) with the old Speedstream modem.

I think that I am going to try to push Bell to switch physical lines. I understand, from a Bell tech, that we have lines coming from quite separate street pedestals, so the performance should be very independent. Your point about water problems is well taken. Historically, we have had water coming up through the floor where the cable enters the building! I seem to have a good contact through DSLReports BellDirect, so maybe he can help me to do this.

I have changed all the cables and the UPS for new ones and the voice PBX equipment has been removed from the phone line.

I wish that I had the option of using cable. Unfortunately, the building is not served by Rogers and I am told that there are no plans, at least in the coming year, to do this. I currently have a backup cellular modem on a minimal plan, but this is expensive for regular use. We are a volunteer community theatre and require reliable service to do ticketing. Fortunately, the summer is a bit of a down time, so I have some "breathing space" to get the disconnect problem fixed.

Thank you
Gordon

 

[Gordon1M]

I am still fighting this disconnect problem. Bell has not been much help ... they have checked the line several times and changed the line profile but that hasn't made much of a difference. I am now running with just one computer connected directly to a HH1000 LAN port (no router and the HH1000 hub negotiates the PPPoE connection) . Disconnects are still happening but these seem to be less frequent than when the RT-N66U negotiates the connection. Are there any parameters set in the N66U, which might be different to those used by the HH1000, e.g. number of attempts to obtain a LCP echo response, timeout parameter etc? The following is a typical HH1000 log entry when a disconnect happens.

28.07.2015 04:15:24 WRN PPPOE Connection to the Internet has been terminated.(Reboot,Reconfiguration,forced termination)
28.07.2015 04:15:29 ERR PPPOE Connection has been terminated: remote station is not answering to LCP Echo requests.
28.07.2015 04:15:30 INF SYS TR-069 connectivity to (postact.cwmp.acs.bell.ca) has failed
28.07.2015 04:15:30 ERR PPPOE PPPoE error: timeout
28.07.2015 04:15:32 ERR PPPOE PPPoE error: timeout
28.07.2015 04:15:37 ERR PPPOE PPPoE error: timeout
28.07.2015 04:16:15 ERR PPPOE PPPoE error: timeout
28.07.2015 04:16:16 ERR PPPOE PPPoE error: timeout
28.07.2015 04:16:18 ERR PPPOE PPPoE error: timeout
28.07.2015 04:16:22 ERR PPPOE PPPoE error: timeout
28.07.2015 04:16:26 INF XDSL ADSL connectivity is up
28.07.2015 04:17:00 ERR PPPOE PPPoE error: timeout
28.07.2015 04:17:01 ERR PPPOE PPPoE error: timeout
28.07.2015 04:17:03 INF PPPOE PPPoE connection successfully established with user credentials.
28.07.2015 04:17:12 INF SYS TR-069 connectivity to (postact.cwmp.acs.bell.ca) has been initiated
28.07.2015 04:17:15 INF SYS TR-069 connectivity to (postact.cwmp.acs.bell.ca) has been closed

I am wondering if the N66U might be less tolerant when the above types of errors occur.

Thank you
Gordon

 

[Brett Lockerbie]

I am still fighting this disconnect problem. Bell has not been much help ... they have checked the line several times and changed the line profile but that hasn't made much of a difference. I am now running with just one computer connected directly to a HH1000 LAN port (no router and the HH1000 hub negotiates the PPPoE connection) . Disconnects are still happening but these seem to be less frequent than when the RT-N66U negotiates the connection. Are there any parameters set in the N66U, which might be different to those used by the HH1000, e.g. number of attempts to obtain a LCP echo response, timeout parameter etc? The following is a typical HH1000 log entry when a disconnect happens.

28.07.2015 04:15:24 WRN PPPOE Connection to the Internet has been terminated.(Reboot,Reconfiguration,forced termination)
28.07.2015 04:15:29 ERR PPPOE Connection has been terminated: remote station is not answering to LCP Echo requests.
28.07.2015 04:15:30 INF SYS TR-069 connectivity to (postact.cwmp.acs.bell.ca) has failed
28.07.2015 04:15:30 ERR PPPOE PPPoE error: timeout
28.07.2015 04:15:32 ERR PPPOE PPPoE error: timeout
28.07.2015 04:15:37 ERR PPPOE PPPoE error: timeout
28.07.2015 04:16:15 ERR PPPOE PPPoE error: timeout
28.07.2015 04:16:16 ERR PPPOE PPPoE error: timeout
28.07.2015 04:16:18 ERR PPPOE PPPoE error: timeout
28.07.2015 04:16:22 ERR PPPOE PPPoE error: timeout
28.07.2015 04:16:26 INF XDSL ADSL connectivity is up
28.07.2015 04:17:00 ERR PPPOE PPPoE error: timeout
28.07.2015 04:17:01 ERR PPPOE PPPoE error: timeout
28.07.2015 04:17:03 INF PPPOE PPPoE connection successfully established with user credentials.
28.07.2015 04:17:12 INF SYS TR-069 connectivity to (postact.cwmp.acs.bell.ca) has been initiated
28.07.2015 04:17:15 INF SYS TR-069 connectivity to (postact.cwmp.acs.bell.ca) has been closed

I am wondering if the N66U might be less tolerant when the above types of errors occur.

Thank you
Gordon

Hello Gordon...did you find any resolve for your LCP Echo Disconnects? I have had this problem with my Bell HomeHub 1000 for months. Very intermittant, frustrating and difficult to trouble-shoot. ( I am not using any additional Router). Any thoughts would be appreciated. Thank you.

 

[heysoundude]

Bell is and has been pushing Fibe in my neighbourhood for a while now. Part of this is because they don't see the point of upkeep on the copper/analog lines any longer. That's an indicator to me to move to cable until bell is forced to bring fibre to your house at no cost to you. Totally not the answer you're likely looking for, but if their line/modem is the problem...you know?
I'm looking at teksavvy for their 120/10 cable package (and goodbye cogeco cable tv).

 

[Brett Lockerbie]

Bell is and has been pushing Fibe in my neighbourhood for a while now. Part of this is because they don't see the point of upkeep on the copper/analog lines any longer. That's an indicator to me to move to cable until bell is forced to bring fibre to your house at no cost to you. Totally not the answer you're likely looking for, but if their line/modem is the problem...you know?
I'm looking at teksavvy for their 120/10 cable package (and goodbye cogeco cable tv).

I have certainly considered going with Cable (Cogeco) here as well. However if I do that, I lose my Bell email address that I've had for over twenty years and I don't want to give that up. I'm also a Bell retiree so the discount I get isn't hard to take either. I am in a village of about 3500 and so waiting for 'Fiber to the House', could be years. Thanks for your input though. If it persists, I will at least swap out the HH1000 and see if the trouble continues.