Painfully slow connection; can't get VM to help

[Sephiroth]

Originally Posted by minichaz

Okay, thanks. Is there any way through to tech support other than 151? As I said if I go this route I don't seem to be able to get out of India. I'll try the newsgroups.

Things are actually _much_ better this morning. Speed is usable and packet loss on that hop is only about 2%. Perhaps this is a congestion related issue? I'll monitor it during the day.

Thanks again,
Charlie

Just to pick up on the 10.xx hop. The packet loss there is not reckonable into any problem you have. That address is the privately assigned IP address of the CMTS - the "you" side if you like. Your router or firewall could/would reject return pings. So that hop can be ignored.

[minichaz]

Originally Posted by Sephiroth

Just to pick up on the 10.xx hop. The packet loss there is not reckonable into any problem you have. That address is the privately assigned IP address of the CMTS - the "you" side if you like. Your router or firewall could/would reject return pings. So that hop can be ignored.

Hi Sephiroth. Can you explain what you mean? I'm curious because that packet loss was 40% yesterday when things were bad and less than 2% this morning when things were working properly. Surely if my firewall was blocking the relevant ICMP packets I'd see 100% loss all the time? I'll keep an eye on it but it does appear directly related to the issue.

Surely the "me" side is my cable modem which has a different 10.*.*.* IP address. Isn't the CMTS the start of the VM side?

As I understand it the CMTS won't just be me but will be shared with many others. With students returning for their new term could this be congestion?

Many thanks,
Charlie

[Sephiroth]

Originally Posted by minichaz

Hi Sephiroth. Can you explain what you mean? I'm curious because that packet loss was 40% yesterday when things were bad and less than 2% this morning when things were working properly. Surely if my firewall was blocking the relevant ICMP packets I'd see 100% loss all the time? I'll keep an eye on it but it does appear directly related to the issue.
[SEPH]: The 10.xx address is the "you" side of the UBR/CMTS (a private IP address supplied to the cable modem by the CMTS) and every path ping will likely show some/all packets as discarded. The fact that you get through to the next node is proof that nothing is wrong due to node 2.

The more technical explanation is that your router or modem simply doesn't respond to some or all ICMP requests (i.e. it doesn't create ICMP TTL expired packets). It depends on what the cable modem or router is doing at the time. Or it's the router firewall rejecting pings (but then it'll be 100% all the time).

Surely the "me" side is my cable modem which has a different 10.*.*.* IP address. Isn't the CMTS the start of the VM side?
[SEPH]: Not really. The CMTS will have a 68.xx or 203.xx public (registered) IP address.

The CM must talk to the CMTS. The 10.xx private address is issued to the CM by the CMTS allowing the CM to communicate with the CMTS.

As I understand it the CMTS won't just be me but will be shared with many others. With students returning for their new term could this be congestion?
[SEPH]: That's a different matter. If there is no network issue (a very complex subject) then poor performance is usually down to congestion.

That can be cause by oversubscription, unbalanced CMTS connections, too many homes connected to the street box and/or those students who do their excessive thing.

Many thanks,
Charlie

Fair enough. It's a complex subject.

[minichaz]

Thanks for that. I'll do some more research... At least I will when my connection is usable. Same old story tonight. Incredibly slow and WinMTR showing 40% loss between me and 10.188.240.1 (and of course to all sites on any route beyond). This has to be congestion, right?

Thanks as always.

[Sephiroth]

Originally Posted by minichaz

Thanks for that. I'll do some more research... At least I will when my connection is usable. Same old story tonight. Incredibly slow and WinMTR showing 40% loss between me and 10.188.240.1 (and of course to all sites on any route beyond). This has to be congestion, right?

Thanks as always.

If I could see that in pathping type format I could give you a more informed opinion. In principle, yes.

[Ignitionnet]

Originally Posted by Sephiroth

Just to pick up on the 10.xx hop. The packet loss there is not reckonable into any problem you have. That address is the privately assigned IP address of the CMTS - the "you" side if you like. Your router or firewall could/would reject return pings. So that hop can be ignored.

Packet loss to the 10.x hop is most certainly worth considering - on all CMTS apart from the BSR whose first hop is a public address that is the CMTS.

Seph - a traceroute is simply a ping repeated with incrementing TTL until destination reached. The Cisco kit answers from the 10.x address which the cable modems talk to so that is what is seen. Either way it's the first hop after the HFC network and should be considered within diagnostics.

The cable modem does not show on any traceroute, it bridges all packets apart from those directed to its' management address so doesn't decrement TTL and doesn't appear in traceroutes. One can ping the cable modem's 10.x address if you happen to know it, it will happily respond, but it's a pointless exercise.

[Sephiroth]

[QUOTE=Broadbandings;34945874]Packet loss to the 10.x hop is most certainly worth considering - on all CMTS apart from the BSR whose first hop is a public address that is the CMTS.

Seph - a traceroute is simply a ping repeated with incrementing TTL until destination reached. The Cisco kit answers from the 10.x address which the cable modems talk to so that is what is seen. Either way it's the first hop after the HFC network and should be considered within diagnostics.
[SEPH]: I see the point your making - packet loss TO the 10.xx hop. What most people notice is the packet loss passed FORWARD from the 10.xx hop - and this can be ignored for the reasons I've given.

......QUOTE]

[SuzieQ123]

Hi,

I was wondering if someone could help me, please? For the past week or so my connection has been awful - it cuts off constantly and I simply cannot do anything other than look at sites. Downloading is impossible.

Looking at these numbers, I get the impression something is wrong (at least with the downstream SNR)

Could someone confirm if this is the case please and whether I need an engineer?

Many thanks!

Type : Ambit ETH/USB Combo Cable Modem
Cable modem : Ambit Cable Modem
IP address :
Downstream Receive Power Level : -1.15 dBmv
Downstream SNR : 29.86 dB
Upstream Transmit Power Level : 45.00 dBmv

Maximum Downstream Data Rate : 10242000 bps
Maximum Upstream Data Rate : 514000 bps

I'm in the Nottingham area and an NTL customer (used to be Diamond Cable years ago)

[Sephiroth]

Can we see all the stats please (the modulation data is missing), plus the event log? On the face of it your SNR is too low. First port of call is to check that your coax is firmly screwed in at both ends.

Second, search this forum on the term NOTTINGHAM to see how others are afflicted.

[SuzieQ123]

My apologies. I didn't realise I had missed things out.

Network Access : Enabled
Maximum Downstream Data Rate : 10242000 bps
Maximum Upstream Data Rate : 514000 bps
Maximum Number of CPEs : 1
Software Upgrade File Name : ntlhm100_ntl0003.cpr
Software Upgrade AdminStatus : allowProvisioningUpgrade
Software Upgrade OperStatus : completeFromProvisioning

Downstream Status
Downstream Lock : Locked
Downstream Frequency : 402750000 Hz
Downstream Modulation : 256 QAM
Downstream Receive Power Level : -1.02 dBmv
Downstream SNR : 30.08 dB

Upstream Status
Upstream Frequency : 25800000 Hz
Upstream Transmit Power Level : 45.00 dBmv

I don't appear to have an "event log" though when I log into http://192.168.100.1/

I have checked the coax at both ends and they appear to be firmly attached.

Many thanks for your help

[Peter_]

Originally Posted by SuzieQ123

My apologies. I didn't realise I had missed things out.

Network Access : Enabled
Maximum Downstream Data Rate : 10242000 bps
Maximum Upstream Data Rate : 514000 bps
Maximum Number of CPEs : 1
Software Upgrade File Name : ntlhm100_ntl0003.cpr

I would recommend that you call in and get that old modem swapped as a NTL home 100 will struggle with 10Mb as it is rather an old piece of kit.

Call 151 from a Virgin phone or 0845 454 1111 from any other phone.

[SuzieQ123]

Originally Posted by Moldova

I would recommend that you call in and get that old modem swapped as a NTL home 100 will struggle with 10Mb as it is rather an old piece of kit.

Call 151 from a Virgin phone or 0845 454 1111 from any other phone.

Many thanks, Moldova

[Sephiroth]

If you do call it in get them to look at your SNR. For the 256QAM modulation, you're right on the lower limit without any headroom for marginal conditions.

It may not be your modem that's causing the poor experience.

You can post your stats on the support newsgroup; they'll notice it's an old modem and offer the replacement, but will also test your line if you make the point.

151 might get you onto the offsahore call centre and although they'll happily arrange a new modem, they're not likely to want to understand anything about the low SNR.

[mjderrin]

Hi I just wanted to add that I have been suffering very poor upload and download speeds for some time - seriously bad making working from home a waste of time - quicker to do the 100 mile round trip to the nearest offce.
I contacted tech support and they said I didn't need a new modem - clearly from the discussion here they are wrong - I have the same modem I was given when I joined ntl back in 2003 - I am on an L package and I pay £25 a month for it - everyone else seems to pay less and get more - your discussion has helped me resolve to call CS tomorrow morning - thanks
BTW - the downstream power level is 4.2dBmV, snr 23.8dB; upstream pwr: 46.0dBmV, symbolic rate 2560 k sym/sec

[Sephiroth]

Originally Posted by mjderrin

Hi I just wanted to add that I have been suffering very poor upload and download speeds for some time - seriously bad making working from home a waste of time - quicker to do the 100 mile round trip to the nearest offce.
I contacted tech support and they said I didn't need a new modem - clearly from the discussion here they are wrong - I have the same modem I was given when I joined ntl back in 2003 - I am on an L package and I pay £25 a month for it - everyone else seems to pay less and get more - your discussion has helped me resolve to call CS tomorrow morning - thanks
BTW - the downstream power level is 4.2dBmV, snr 23.8dB; upstream pwr: 46.0dBmV, symbolic rate 2560 k sym/sec

If your downstream modulation is 64QAM then your SNR is at the lower threshold for your modulation and not sustainable. This needs calling in. I suggest you use the support newsgroup for a swift reply and resolution.

If you're on Vista or Windows 7, you would use Windows Live Mail (a standard part of your system). To get on the VM Newsgroup you need to be operating through the VM broadband system.

You'd click the "Newsgroups" link at the bottom of the screen and then Tools/Accounts then Add/Newsgroup Account.

Give your Display Name (like mjderrin), your e-mail address then your server address for the newsgroup which is "news.virginmedia.com".

You'll then be able (via the Newsgroups menu item) to select the Virgin media icon on the left of a search panel, click on "All" and wait a few minutes till it's finished listing thousands of newsgroups.

Scroll down to "Virginmedia.support.broadband.cable" and double click it.

And you're there. You can post your stuff including the evidence you've collected.