Intermittent DSL Connection

[floresca]

The internet connection randomly drops - Sometimes only lasts for 10 min and we have to reboot the modem.

We have just been sent a new modem - 7500 - they upgraded our speed to the 4-7.  The modem is PPPoe mode & we use the 7500 for wireless.  

2 months ago, a tech came out and placed a homerun filter at the NIC.  This solved the problem for about 2-3 weeks and then the issues started back up again.

 
A possible clue:  The dsl light stays solid green but the internet light turns off when we don't have any internet connection.

Thanks for any ideas.

[floresca]

Transceiver Stats

-these were taken when the modem was connected in its normal place.  It was running extremely slow at this point. Many times the transceiver stats show 5886 for speed. 

Transceiver Revision    A2pB020b3.d20h

Vendor ID Code             4D54

Line Mode                       ADSL_re-adsl

Data Path                        FAST
Transceiver Information

                                                 Down Stream Path       Up Stream Path

DSL Speed (Kbits/Sec)       1448                                561

Margin (dB)                            6.4                                   7.2

Line Attenuation (dB)           52.0                                 33.0

Transmit Power (dBm)         0.0                                   11.0

[floresca]

TRANCEIVER STATS taken when modem was connected outside to the NIC 

Transceiver Revision    A2pB020b3.d20h

Vendor ID Code            4D54

Line Mode                      ADSL_2plus

Data Path                        FAST

 Transceiver Information                            Down Stream  Up Stream Path

DSL Speed (Kbits/Sec)                              5886                  861

Margin (dB)                                                   6.6                    12.5

Line Attenuation (dB)                                 44.5                   21.4

Transmit Power (dBm)                               0.0                     12.9

[smith6612]

Have you ever seen the stats looking like the way they do at the NID at the location where the DSL modem tends to sit? It is looking like a problem inside of your home right now, perhaps due to old or damaged wiring.

If you want to try and rule out the issues occurring in your home, the first step I would take would be to check your jacks for any dirt or corrosion. I would also investigate the wiring behind the jacks, ensuring it is tight, but not over-tightened and the connections are all free of corrosion or dirt. I would also double check all DSL filters to ensure they are installed facing the correct way (they work the best in one direction), and that all devices are behind a DSL filter. Failing that, if your NID's downstream attenuation levels as reported by the modem are correct, I would suggest setting up a "Home Run" setup. This would involve running a dedicated pair of CAT3 cabling for the DSL modem itself, but it would allow you to filter out the home wiring which will rule out many issues caused by wiring inside of the home.

[floresca]

We already have a home run filter.  The tech put it in about 2 months ago.

And, yes, we do get stats in the living room that look just like the stats from when the modem is plugged into the NID.

Am I correct that having the home run filter installed in the NID means that the dsl signal is now separate from the voice signal - meaning that:

1.  faulty wiring inside the house can be eliminated as part of the problem?

2.  do we still need filters on each individual phone if we have the home run filter?

Thanks

[floresca]

UPDATE:

A tech came 2 days ago.  She was definite that the home run filter should eliminate any problems coming from house wiring.  Her thought was that our line could not handle the 4-7 mbps service and that if we asked to be optimized at 3 then that would solve our problems.  Called and we are now at 3.

The problem is still occurring, though.  Any ideas?

Here is the system log from the modem.

Time	Severity	Process	Details
Oct 28 12:57:46	daemon.info	cwmpd[338]:	CWMP: Final Http Return Code = 204
Oct 28 12:57:46	daemon.info	cwmpd[338]:	CWMP: Inform (200). The following events are true:4 VALUE CHANGE
Oct 28 12:57:46	daemon.info	cwmpd[338]:	CWMP: Inform Sent Successfully
Oct 28 12:57:43	daemon.err	cwmpd[356]:	cwmpDDNSCallback: Result of DDNS Update (ipupdate_once) ->3
Oct 28 12:57:35	daemon.notice	ip-up[6425]:	ppp0: nameserver:71.252.0.12
Oct 28 12:57:35	daemon.notice	ip-up[6425]:	ppp0: nameserver:71.242.0.12
Oct 28 12:57:35	daemon.notice	ip-up[6425]:	ppp0: default route:ppp0/10.11.13.1
Oct 28 12:57:34	daemon.err	pppoe-relay[1127]:	PADS packet from 00:90:1a:a1:15:96 on interface nas1 does not have Relay-Session-Id tag
Oct 28 12:57:33	daemon.err	pppoe-relay[1127]:	PADO packet from 00:90:1a:a1:15:96 on interface nas1 does not have Relay-Session-Id tag
Oct 28 12:57:33	user.notice	root:	ppp0 started on nas1
Oct 28 12:57:33	user.notice	root:	ppp0 stopped on nas1
Oct 28 12:57:32	user.warn	kernel:	636f6e66 (epc == 2ac2daec,ra == 2abce02c)
Oct 28 12:53:18	user.crit	kernel:	ADSL link up,interleaved,us=445,ds=1788
Oct 28 12:53:18	user.crit	kernel:	ADSL G.992 message exchange
Oct 28 12:53:14	user.crit	kernel:	ADSL G.992 channel analysis
Oct 28 12:53:09	user.crit	kernel:	ADSL G.992 started
Oct 28 12:53:02	user.crit	kernel:	ADSL G.994 training
Oct 28 12:52:57	user.crit	kernel:	ADSL link down
Oct 28 12:52:56	user.crit	kernel:	ADSL G.992 message exchange
Oct 28 12:52:52	user.crit	kernel:	ADSL G.992 channel analysis
Oct 28 12:52:52	daemon.err	cwmpd[356]:	ddnsWanIPCallback WAN State Down
Oct 28 12:52:47	user.crit	kernel:	ADSL G.992 started
Oct 28 12:52:45	daemon.err	cwmpd[356]:	ddnsWanIPCallback WAN State Down
Oct 28 12:52:41	user.crit	kernel:	ADSL G.994 training
Oct 28 12:52:40	daemon.notice	ip-down[6040]:	ppp0: 89479 bytes sent,38358 bytes received
Oct 28 12:52:39	user.crit	kernel:	ADSL link down

[crnfords]

I have had a similar problem since at least July this year, several tech visits to disprove it was my equipment and several server changes by Verizon tech support have not helped. It appears that "someone" is disconnecting my connection to the inet, poss a hardware tech, but more likely a software dude. I have tracked the service lapses and see most of the outages start just before or at the lunch hour and/or close of the business day. The and/or is because sometimes it happens twice a day, but never on a Sunday or holiday. I have the modem logs to prove these statements. I dump the log content to a word file before every reboot as the modem reboot clears the logs.

I am in the 23693 zip code area, pls reply if you have similar issues or on my zip area and are having any outage issues.

[crnfords]

Happened again today, fourth consecutive day this week, 1107 AM and 0345 PM. Still no return call from any tech support persons that are supposed to be working this issue.

[smith6612]

At your distance, 5Mbps should be doable with ADSL2+, but FastPath would ideally need to be taken off of the line for stability reasons. If the statistics are close to each other in the living room as they are in the NID, then the wiring between the splitter and the jack is just fine. The Home Run should isolate any issues inside of your home, and will in fact eliminate the need for DSL filters in your home.

It's hard to say what Verizon had your line set to speed-wise without seeing more of a log, but it looks like the drop may have been from Verizon dropping you all the way down to 1.5Mbps, which compared to 5Mbps, is awful-slow. If you're still having drops even at 1.5Mbps, or are dropping at 3Mbps, the problem is not with the speed being too high and Verizon should technically set your speed back to where it was (3-5Mbps range) and solve the problem while it's under a condition where it is most prone to happening.

[crnfords]

This is a slice of the modem log after rebooting from today's 11AM daily outage, show we are using same name servers and are having very similar problems with internet connection:

Nov 5 10:09:59	daemon.info	dnsmasq[325]:	using nameserver 71.252.0.12#53
Nov 5 10:09:59	daemon.info	dnsmasq[325]:	using nameserver 71.242.0.12#53
Nov 5 10:09:59	daemon.info	dnsmasq[325]:	reading /var/etc/resolv.conf
Nov 5 10:09:51	daemon.info	cwmpd[338]:	CWMP: SSL Status = SSL Connection Succeeded,TCP Status = TCP Connected,Authentication Type = Basic
Nov 5 10:09:47	daemon.notice	net_mgr[358]:	RIPConfig: no interfaces are enabled
Jan 1 00:01:18	daemon.notice	net_mgr[358]:	Routes transitioning to Up state on ppp0.
Jan 1 00:01:16	daemon.notice	net_mgr[358]:	WAN came up before SIPALG was configured
Jan 1 00:01:15	daemon.notice	net_mgr[358]:	QoS on ppp0 transitioning to Up state.
Jan 1 00:01:10	daemon.notice	ip-up[1778]:	ppp0: connected:local={edited for privacy},remote=10.13.1.1
Jan 1 00:01:10	daemon.notice	pppd[1763]:	Westell PPP Environment:device:nas1 ifname:ppp0 linkname:config0 iplocal:{edited for privacy} ipremote:10.13.1.1 ppplogname:root orig_uid:0 usepeerdns:1 dns1:71.252.0.12 dns2:71.242.0.12
Jan 1 00:01:10	daemon.warn	pppd[1763]:	pppd[1763] caused InternetUp light to be lit
Jan 1 00:01:10	daemon.notice	pppd[1763]:	Westell ip_up()