Internet Disconnects after a few hours usuage

[rdhw]

Originally Posted by mr-b

System up time = 1 days 00h 12m 16.00s
Ethernet single-collision errs.1 = 4731
Ethernet multiple-collision errs.1 = 3616
Ethernet deferred xmits.1 = 43865

That's a high ethernet error rate for just one day. What model of router do you have? What spec does its WAN port have? 10Mbps or 100Mbps?

Upstream transmit signal power = 56.0 dBmV

Not good - that is pushing right at the limit. You might need an NTL techie visit to sort out your upstream power

Docsdiag sometimes doesn't work as somehow I can't always access 192.168.100.1 behind my router.

That's a problem with the Ambit modems and routers: the Ambits sometimes reply from a different IP address than 192.168.1.100, so routers drop the replies.

[mr-b]

Originally Posted by rdhw

That's a high ethernet error rate for just one day. What model of router do you have? What spec does its WAN port have? 10Mbps or 100Mbps?Not good - that is pushing right at the limit. You might need an NTL techie visit to sort out your upstream powerThat's a problem with the Ambit modems and routers: the Ambits sometimes reply from a different IP address than 192.168.1.100, so routers drop the replies.

Tx Robin for the info.
The router is a Netgear RP114 with a 10M WAN port. Rather elderly but it's been reliable up to now. The only time I've ever had to reboot it was when a P2P app seemed to swamp it with connections and it started dropping UDP packets e.g. DNS lookups stopped working.

I think I will ask for an engineer visit now this issue has happened twice in the past week.
That Ambit CM bug is a bit silly though. Might be an idea to add it to your page.

Originally Posted by mr-b

Tx Robin for the info.
The router is a Netgear RP114 with a 10M WAN port. Rather elderly but it's been reliable up to now. The only time I've ever had to reboot it was when a P2P app seemed to swamp it with connections and it started dropping UDP packets e.g. DNS lookups stopped working.

I think I will ask for an engineer visit now this issue has happened twice in the past week.
That Ambit CM bug is a bit silly though. Might be an idea to add it to your page.

I wouldnt generally say its a bug, it could be down to our internet download/uploading activity or a bad configuration

[rdhw]

Originally Posted by mr-b

The router is a Netgear RP114 with a 10M WAN port.

For the optimal performance, if your router has a fixed 10Mbps WAN port, then you should configure the Ambit's ethernet socket also to be 10Mbps half-duplex, rather than auto-detect. You can do this in the Ambit configuration pages.
__________________

Originally Posted by xathras

I wouldnt generally say its a bug

If you send a request packet to 192.168.1.100, and the reply is tagged as coming from a totally different IP address (the ISP-side IP address of the CM), then it's a real bug, and a bad one.

[mr-b]

Originally Posted by rdhw

For the optimal performance, if your router has a fixed 10Mbps WAN port, then you should configure the Ambit's ethernet socket also to be 10Mbps half-duplex, rather than auto-detect. You can do this in the Ambit configuration pages.

How do you do this? I couldn't find any way to configure this from the Ambit web page. The only things that I can apparently change are the password and Set Search Frequency Parameters.

[carlingman]

Originally Posted by mr-b

How do you do this? I couldn't find any way to configure this from the Ambit web page. The only things that I can apparently change are the password and Set Search Frequency Parameters.

Without the router in play and connecting via the 192.168.100.1 page rather than through docs diag.

From the first page of the diagnostics you should see an option to set ethernet type and a drop down box to choose this setting.

[mr-b]

Originally Posted by carlingman

Without the router in play and connecting via the 192.168.100.1 page rather than through docs diag.

From the first page of the diagnostics you should see an option to set ethernet type and a drop down box to choose this setting.

Docsdiag is for diag - not mgmt.
There is nowhere at all on the Ambit ntl home 200 modem http:/192.168.100.1 web page to set the ethernet port speed that I can see. Can you be more specific?

[ian@huth]

Doesn't it have "set ethernet type" at the top of the page?

[carlingman]

Yep thats where it is on all the diagnostics pages I have seen ????

[mr-b]

Originally Posted by ian@huth

Doesn't it have "set ethernet type" at the top of the page?

Nope - near the top it has links for "Cable Modem Information", "Set Seach Frequency Parameters" and "Change Password". On the left it has links for Configuration, Status, Operations Configuration, Interface Rate and Event Log. I see the same page with both Firefox and IE6.

But all that pales into insignificance against all the grief I've just had with support.
After another connection failure, they sent an engineer out, who at first cancelled the call because I have a completely separate ntl business telephone line at the same address. After they sent him out again, he adjusted something so that my upstream power is now looking like this:

DOCSIS 1.0 Cable Modem 60678EU <<HW_REV: 1.15 ; VENDOR: Ambit; BOOTR: 3.13.2 ; S
W_REV: 2.67.1011 ; MODEL: 60678EU >>

System up time = 0 days 00h 04m 28.00s
Ethernet deferred xmits.1 = 99
Downstream channel ID = 0
Downstream channel frequency = 402750000 Hz
Downstream received signal power = -5.1 dBmV
Upstream channel ID = 3
Upstream channel frequency = 25584000 Hz
QoS max upstream bandwidth = 300000 bps
QoS max downstream bandwidth = 3072000 bps
SigQu: Signal to Noise Ratio = 34.8 dB
Cable modem status = Operational
Upstream transmit signal power = 51.0 dBmV
Date and Time = 2005-03-18,23:22:38.0
Configuration filename = cmreg-ntlhm200-high.cm

But today it failed again - no DHCP address was obtainable.
I called support, who then insisted that there was nothing wrong with the CM (since my SYNC and RDY lights were on solid) and that the fault must be with my router and that I'd just have to keep rebooting my CM, even though I'd disconnected my router on previous occasions and managed to replicate the exact same IP loss issue with my PC.

I then asked to speak with a supervisor - whom I got to agree that rebooting a CM every few days was not "normal" behaviour, espcially since I'd had mine for 5 years and never experienced this before.

He then advised me that it might have something to do with the fact that there may not be enough IP addresses in my area, but that there was little they could do about it unless there were a lot of other complaints...
If this goes on much longer I can see an ADSL order looming...

[mr-b]

Lost my IP address again today. CS had been due to call me back since it failed yesterday too - but of course they never did.
When I called again, this time their excuse for the constant loss of IP address and necessary CM reboots was that I should be turning the CM off regularly anway, as otherwise it would generate lots of heat!

After I confirmed that they had carried out all their diagnostics I rebooted the CM so it would give me an IP address again - although CS refused to acknowledge that there was a problem, saying that it was the fault f my PC and that I needed to run the cmd:
> netsh interface ip reset resetlog.txt" each time. They didn't know how to do this on a router, but they didn't support routers anyway...

Then after I talked again to a supervisor, he said that I would have to call again when it stopped working again.
I insisted then that they send out an engineer to examine the CM since that is the only thing that has not been replaced yet - so we'll have to see how it goes tomorrow...
__________________

Quick addendum to this:

I also discovered whilst monitoring the DHCP lease behaviour (for just my PC behind the CM) that sometimes the lease is for an hour, and then after another refresh it goes to 2 days! Anyone else seen this?

[Peter_Stanley]

Originally Posted by mr-b

Over the past few months I have started to experience strange CM "disconnection" problems .... the SNR/upstream power levels look fine as per Robin walker's page ... the Sync/Rdy lights are on but there is no internet connectivity... if I reboot the CM everything works fine again.

I've been having a very similar problem for the last couple of weeks. Essentially my connection is working perfectly and then intermittently drops. The cable modem still show solid SYNC and RDY lights, but webpages time out or cannot be found and email servers are not reachable. The problem can last for a few seconds or 20+ minutes and then self-rectify.

Talking to NTL on this is VERY frustrating:

Call 1 - Remove your router - that is causing the problem
Call 2 - (Direct connection from CM to PC) - There's definately a problem, Tech Support intermittantly getting 100% loss (not sure what this means?) when interrogating the CM.
Engineer booked to switch out the CM, but actually just changes the power lead/transformer and adds an attenuator to adjust the signal power levels.
Call 3 - Remote test of the connection with PC off shows 0% loss, with PC on this dropped to 100%, "so it must be the PC". (Well, maybe but it is an intermittent connection problem and we certainly haven't done enough testing to conclude this yet!) Next told to disable ZoneAlarm and NAV (which I refuse to do - have you seen the infection rate statistics for unprotected broadband connections?!). Disable ZA and now we have a 10% loss. (Hmm seems to be fluctuating a lot). Follows a confusing conversation where Tech Support trys to tell me that Ad-Aware SE and Spybot S&D are spyware/adware and to disable them because they're maxing out my connection, even though I state that (a) they're not spyware and (b) they're not even running! Finally tech support states that it must be ZoneAlarm hammering my upstream and suggests I contact ZoneLabs - funny how it isn't affecting anyone else then! Just thought - if the problem is spyware, wouldn't the U/S light be continually lit when the problem occurs? (It isn't).

SO, to test the hypothesis that my PC is the problem, I'm typing this on a different PC that doesn't use ZA or NAV. I'm willing to bet a considerable sum that the problem still occurs...!

[I'm not trying to malign the good NTL tech guys, but there are some who talk out of their backsides IMHO.]

Anyone have any suggestions on what the problem might be?

Just for info, here are the SNR levels etc which look okay I think (need to check Robin Walker's site):

Downstream receive power level: fluctuating between -1.84 and -2.01dBmv
Downstream SNR: fluctuating between 30.25 and 31.00 dB
Upstream transmit power level: 40.50dBmv

The error log shows R02.0 No Ranging Response received, T3 time-out with the last occurence about the same time as I last lost the connection.

Any help much appreciated!

[mr-b]

I empathise completely with your frustrations since your experience sounds very familiar e.g. WTF is "loss"???
Some ppl in support these days really do seem to be paid on how quickly they can fob you off with any old excuse i.e. it must be your [insert random h/w or s/w element].
I've had to ask to speak to a supervisor on multiple occasions to try and get get past all this BS.

Did you check your DHCP lease details?
Also note that the Ambit modem event log can be rather confusing since it's not in date order.

I was lucky though in that the Sync LED on my modem decided to fail completely and so they had to come out and swap my modem (replacement was the same Ambit 200 model). Since then I've had no further trouble - but it is early days ...

[th'engineer]

Originally Posted by Peter_Stanley

Call 3 - Remote test of the connection with PC off shows 0% loss, with PC on this dropped to 100%, "so it must be the PC". (Well, maybe but it is an intermittent connection problem and we certainly haven't done enough testing to conclude this yet!) Next told to disable ZoneAlarm and NAV (which I refuse to do - have you seen the infection rate statistics for unprotected broadband connections?!). Disable ZA and now we have a 10% loss. (Hmm seems to be fluctuating a lot). Follows a confusing conversation where Tech Support trys to tell me that Ad-Aware SE and Spybot S&D are spyware/adware and to disable them because they're maxing out my connection, even though I state that (a) they're not spyware and (b) they're not even running! Finally tech support states that it must be ZoneAlarm hammering my upstream and suggests I contact ZoneLabs - funny how it isn't affecting anyone else then! Just thought - if the problem is spyware, wouldn't the U/S light be continually lit when the problem occurs? (It isn't).

I see some people in tech support have not changed still anthing but the NTL equipment.

I just wonder why we are all having problems after the upgrade disconnects slow speeds etc .

Its like when they had to turn proxies off because of a major problem but not telling anyone.

Or even something was not tested prior to the upgrade, could it be possible that it was introduced without the right technical involvement.
Put in to a date and time unrealistically with errors in a part of the system.

[OldGiraffe]

I have just sent this to as many media contacts as I could find: -

†ωÃà ‚¢Ã¢â‚¬Å¡Ã‚¬Ã‚ÂÂÂ� Can anyone beat this record for poor broadband service given by NTL?
They claim 'always on' or as some would say 24/7. Well, I can accept that there will be the odd occasion when this is impractical but an average of only 10 hrs in 24 for 8 days at best (and still counting) takes the biscuit! One day there was a connexion for only three hours and another day zero. Then there were the four engineer visits when each one knew nothing of the other's findings and the dozen or so calls to telephone support where again each 'agent' knew precious little about what was said to and by other support 'agents' and they knew even less about what the engineers had found, or more often not found. Then there are the network engineers in their own mysterious world divorced from the reality of the field engineers and hardly ever spoken to by the telephone support staff! Never the same engineer calls, never the same person on the telephone and no single reference point for the fault in hand. To add insult to injury each time a telephone call is made one has to go through the routine: listen to the music, press this number, press that number half a dozen times, listen to messages telling us how to suck eggs and then when one gets to talk to a real person, listen to the multitude of different 'solutions' for the one problem, many being absurd. In addition, the automatically generated emails which follow each telephone call with suggestions which are mostly irrelevant. Why not allocate a fault number and give us an early 'press this option' short cut straight to someone with access to the total fault report history?
Come on NTL, let's have some 'joined up' technical support then maybe I will get my broadband back! †ωÃà ‚¢Ã¢â‚¬Å¡Ã‚¬Ã‚ÂÂÂ�

I am still far from happy with your telephone response to my letter 1st of January 2005, to which I asked for a written reply but did not receive one. I will be writing again about my other complaints in that letter.

Meanwhile; please tell me how I can escalate my complaints up the management chain.

NO REPLY FROM NTL!!!!