30M Apparent unknown problem - Bexleyheath area. [Archive]

Adam_swaine

05-03-2012, 00:44

Hey, sorry that i'm cross linking to the official VM forums but the forum codes are different and a simple copy and paste wouldn't work.

So here is my thread

http://community.virginmedia.com/t5/Up-to-30Mb-broadband/Help-me-find-the-answers-anyone-can-post-please/td-p/1076289

Please have a read and hopefully you might be able to enlighten me to what my problem is.

craigj2k12

05-03-2012, 01:13

looks like something LAN side with the jump in base latency, do you use any p2p software?

Adam_swaine

05-03-2012, 01:20

looks like something LAN side with the jump in base latency, do you use any p2p software?

It's not something I'm doing I can guarantee that, in fact I've made sure not do do anything to effect my speeds since the first night of issues.

The last few nights I have literally sat there refreshing the graph at around 7 and watching it go up while I do nothing else, then refreshing at 12 and watching it going down till it goes back to normal at 12:30.

It's an odd one as It appeared out of no where after perfect non interrupted speeds for a good 2-3 weeks.

I'll stress again, definitely not something I'm going to effect the line, I promise.

Thanks for the speedy reply though, it's appreciated. . keep it coming as tech support had nothing to offer and I'm waiting on a VM mod to reply to my other thread on the forums.

Peter_

05-03-2012, 06:39

Have you tried your connection with a friends laptop to rule out your own machine.

Andrewcrawford23

05-03-2012, 07:30

with that high a acket drop i rather doubt it be totalyl down to viign

open comamdn prompt and type netstat -b and netstat -a an pos the resutls here

qasdfdsaq

05-03-2012, 10:38

Seen quite a lot of these reports of bad performance in the last couple weeks - more so than before. Also seen a few VM core network nodes showing unusual signs of congestion. Could it be VM are doing major network upgrade work?

---------- Post added at 10:38 ---------- Previous post was at 10:36 ----------

looks like something LAN side with the jump in base latency, do you use any p2p software?
Thanks to the monstrous buffer bloat VM force on it's users, it's actually quite easy to tell that that isn't the case.

Andrewcrawford23

05-03-2012, 10:38

Seen quite a lot of these reports of bad performance in the last couple weeks - more so than before. Also seen a few VM core network nodes showing unusual signs of congestion. Could it be VM are doing major network upgrade network?

if they are my area is prove of the work acutalyl making bg changes my thinkbronadband monitor now only shows 40ms as the hgiest ping wherese befor eit was 120ms

qasdfdsaq

05-03-2012, 11:07

Heh, my BT line would show a maximum of 20 if they hadn't added 12ms of artificial delay to my service...

Adam_swaine

05-03-2012, 13:19

Tried on 3 different machines.

Cmon, look at my graphs and the graphs at theinternetprovider (as posted in my thread) you can see the correlation between the two, I can guarantee without doubt it isn't on my end.

But here I go

https://www.cableforum.co.uk/images/local/2012/03/67.png
https://www.cableforum.co.uk/images/local/2012/03/68.png

Andrewcrawford23

05-03-2012, 14:21

Adam_swaine

05-03-2012, 14:23

no nothing popping out suspiocuios

can you do me a favour when you are getting the high packet lost form teh monitor can you run these command

ping -n 100 bbc.co.uk
tracert bbc.co.uk
pathping bbc.co.uk

something is strange i needa bit mroe information to try he,p pinpoint it, i think it mostly virign problem

but i think you have shub and it mgith that is faulty or flood procttion needs turned off
Yer i do have a shub but all the un wanted services are turned off.

I'll run those in command for you at around 7:30 when the problem happens, so come back then and we're see what we got.

Andrewcrawford23

05-03-2012, 14:25

its possible the shub is faulty and isnt replying to the pigns it is overlaoded somehow but lets see what those command bring im genereally off and on until 9pmish

craigj2k12

05-03-2012, 14:39

tracert to pingbox1.thinkbroadband.com - thats the address of the TBB ping monitor

Adam_swaine

05-03-2012, 16:19

tracert to pingbox1.thinkbroadband.com - thats the address of the TBB ping monitor

You have to remember it's only 4:20pm, the issue occurs between 7pm and 12:30am, I don't think it's an issue between me and thinkbroadband as it's not just a case of the graph showing latency increase and me not seeing any effects I'm losing download speed and getting increased ping because of it.

But i'll post as you asked.
https://www.cableforum.co.uk/images/local/2012/03/65.png

Andrewcrawford23

05-03-2012, 16:43

You have to remember it's only 4:20pm, the issue occurs between 7pm and 12:30am, I don't think it's an issue between me and thinkbroadband as it's not just a case of the graph showing latency increase and me not seeing any effects I'm losing download speed and getting increased ping because of it.

But i'll post as you asked.
https://www.cableforum.co.uk/images/local/2012/03/65.png

acutally craig gota point can you do the tests i saida above, since the problem exists after 7pm if you do it jsu tnow we get a comparsion off now and then, and yeah go all 3 tests to that address craig gave you, as sicne it pinging you it might help show where the problem is coming form

Adam_swaine

05-03-2012, 16:49

Ok i'll do ping, tracert and pathping. I assume those are the 3 tests you mean.

Andrewcrawford23

05-03-2012, 16:58

Ok i'll do ping, tracert and pathping. I assume those are the 3 tests you mean.

yeah :) pathpign will takes about 3-5 minutes ona normal ok conenction so dnt be surprised if ti takea a while

Adam_swaine

05-03-2012, 17:10

https://www.cableforum.co.uk/images/local/2012/03/63.png

https://www.cableforum.co.uk/images/local/2012/03/64.png

And tracert is on previous page.

Looks like tomorrow should be my slot for a VM staff member to reply to my thread, will see what they say.

Andrewcrawford23

05-03-2012, 17:17

when you run the ping command after 7pm make sure it is ping -n 100 pingbox1.thinkbroadband.com the -n 100 is crucially as it gives it chance to sho packet lost but i dnt think your getting any just now as the pings are low

qasdfdsaq

05-03-2012, 17:35

Good time for me to beta test my new qasi-mon service methinks.

Adam_swaine

05-03-2012, 17:43

Good time for me to beta test my new qasi-mon service methinks.

Oh no, this sounds mischievous.

qasdfdsaq

05-03-2012, 17:47

Yes, a connection monitor is really really mischievious.

Adam_swaine

05-03-2012, 17:56

Yes, a connection monitor is really really mischievious.
:jk:

I appreciate all the input from your lot btw, i'll come back at 7 with those stats.

Adam_swaine

05-03-2012, 20:15

https://www.cableforum.co.uk/images/local/2012/03/59.png
https://www.cableforum.co.uk/images/local/2012/03/60.png

https://www.cableforum.co.uk/images/local/2012/03/61.png

https://www.cableforum.co.uk/images/local/2012/03/62.png

Andrewcrawford23

05-03-2012, 20:17

seems the problem is happening at the core network and by the looks of it before ti goes onto linx so i say it is congestion there

Adam_swaine

05-03-2012, 20:25

Andrewcrawford23

05-03-2012, 20:30

I'm guessing basic tech support on the phone aren't able to inform me about that then? As both times they have said said the utilisation in my area is nothing of concern :/ Would i need to get put forward onto someone higher up in order to truely find out?

Or is congestion different to utilisation

congestion in the form i mean is different

utiliastion is my books is the perecent of the ubr capicty is use, i can say for sure if it is over or under ulisated if tech supoprt says it aint then ill take the word on it

congestion in my books is to do with teh backbone is the links from virign to the outside world internet, tech support cant see that i could but could since i know ho to find it

basically lets says they have 10gbit pipe from the core to linx it is congested ie full due to evefryone using the internet but your local ubr isnt

qasdfdsaq

05-03-2012, 20:30

No, the congestion is at the local node. The core network is fine.

Andrewcrawford23

05-03-2012, 20:32

No, the congestion is at the local node. The core network is fine.

as i said i mean sometihng diffferent with congestion look at the pings at the local node there lowest, it the hop jsut before linx that has packet lost and high pign that sugges thte problem is there

Adam_swaine

05-03-2012, 20:40

Andrewcrawford23

05-03-2012, 20:43

So can anything be done on my part in getting this seen to? As atm they aren't going to raise the issue.

What really has me thinking is, from the 11th-29th, perfect service. . . full speeds every day no spikes.

it possible there doign work in yoru area and they might have moved a low more people onto where you conect to whislt hte workign is getitng done. i cant be sure though

there not really much you can do to get it raised if ia m right it at pipe goign to linx virign eventalyl will upgrade it but majority of the tech support wouldnt be able to find it the bes ti can suggest is post those information on community forum hopefully second line can spot it and get ti raised

Adam_swaine

05-03-2012, 20:49

qasdfdsaq

05-03-2012, 20:51

Hmm, well whatever it is, it's not showing up at my end (either of my four ends). Looks like bad routing. That one particular link/hop (if accurate) is badly congested while there are several alternative paths with no such load problems.

Adam_swaine

05-03-2012, 20:55

Yer thats the problem, it's not showing up so 1st line support can't really do anything.

And I take it by link/hop you mean http://www.ukinternetreport.co.uk/cmts/details.php?area=bexl&town=Bexleyheath If you do mean that, is there a chance to get onto another link/hop by complaining constantly.

Andrewcrawford23

05-03-2012, 20:56

Yer I've got two threads on the VM forums, one for everyone to post in and one that i'm currently waiting a reply from a VM admin.

Do you advise to keep doing speed test before and after, as well as pathping/ping/tracert so i've got proof incase this issue progresses?

Also is there no one at all I can ring to find out if it is upgrade work? As I've still got a few days till VM respond.

if the tech support agent is good they will know where to look but most dnt so unlikely, the other problem is it might be anotehr part of the ubr getting work done and they might onlysee hte bit you conenct to and as such might nto find work getting done to the bit your on but work doine elsewhere

---------- Post added at 20:56 ---------- Previous post was at 20:55 ----------

Yer thats the problem, it's not showing up so 1st line support can't really do anything.

And I take it by link/hop you mean http://www.ukinternetreport.co.uk/cmts/details.php?area=bexl&town=Bexleyheath If you do mean that, is there a chance to get onto another link/hop by complaining constantly.

not as such again 1st lien are probally not goign to be to help if it is rotuing which i think qas is right it is routing issue as if it took another route it be better

qasdfdsaq

05-03-2012, 21:03

All of my four "different" routes in look fine. Adam, try changing your IP address. Looks like you're not connected through a PooperHub so forcing an IP change should be easy enough. Changing your IP should change your routing. That said, there's local node congestion too which affects everything, unlike bad routing which only affects traffic to/from certain places.

Adam_swaine

05-03-2012, 22:44

Just looking at my bill, I have a £10 deducted from my next bill applied on the 2nd. Would that be related?

---------- Post added at 22:44 ---------- Previous post was at 21:03 ----------

Just spoke to tech for a good hour while we went through more checks, he took control of my pc and did what you guys told me to do (ping, pathping, tracert), I also showed him my thinkbroadband monitor along with the internetreport.co.uk/cmts which supports the fact my issues are real and not just thinkbroadband showing false latency increase.

After speaking to second line support he said they think it might be to do with

http://desmond.imageshack.us/Himg838/scaled.php?server=838&filename=49b7dfe460944ec52dcbdb9.png&res=medium

Which apparently is the modem part of the superhub and has got a tech to come out to me tomorrow between 12-4. I'm hoping the tech can pin point the issue, but I don't know because he will be coming round at a time where the issue isn't happening.

Also I don't get how an issue with the modem part of the hub could make my latency increase at the same time every night and decrease at the same time, but hey hoo at least i'm getting a tech out.

qasdfdsaq

06-03-2012, 00:09

No, that's the internal (DOCSIS side) IP of the CMTS/UBR.

Nothing to do with the modem or Superhub. It's in the VM "exchange" and is perfectly normal.

Adam_swaine

06-03-2012, 00:51

Ah, you seem to know your stuff, appreciate all the input. I guess i'll probably end up getting a new hub tomorrow, i'll try and explain to him what you guys have told me.

You mentioned changing my IP, I'm on a superhub, so what would be the easiest way to do that.

---------- Post added at 00:51 ---------- Previous post was at 00:32 ----------

Also what should i say to the tech coming tomorrow, I don't want him to come round and simply replace my hub (would laugh if this was the issue) It would be good if I could translate something from this forum to him to hopefully give him a better understanding of what it could be.

Adam_swaine

06-03-2012, 12:11

Anyone? He should be here from now, will be great if I can make the most of him coming round.

Andrewcrawford23

06-03-2012, 12:26

Anyone? He should be here from now, will be great if I can make the most of him coming round.

it be hard to put it to the engineer as qas has said there will be congestion at the local node as well as otehr things effecting it, the best you coudl do is get the tech to try get hte power levles for downstream closer to 0 and upstream about 40 and snr about 40 not sure there much else you can get hte tech to do as it most liekly netwokrs will need to do work you could get him to raise the issue to network to look at it

Adam_swaine

06-03-2012, 12:32

Thanks, yer I guess him coming round and swapping the superhub and attenuator will at least rule it out as being a hardware issue.

Adam_swaine

06-03-2012, 15:07

So.....swapped the hub and attenuator like I guessed. . . only time will tell, actually it won't I know the hardware isn't the issue :/

He did say they have had a lot of people reporting similar issues but they havn't been told anything as to why its happening, even though a lot of them think it is due to upgrade work.

I'll just keep taking records and calling and try and get some money off for the bad evening service.

Andrewcrawford23

06-03-2012, 15:09

good luck with refuns :) i hope it works out fien nwo for you hopefulyl i have bene of some help

Adam_swaine

06-03-2012, 15:15

Yer you've all been of help, i'll post back when normal service resumes as to be honest. . . there is probably nothing more to it then someone doing some upgrade work and that can't be stopped unfortunately.

It would just be great for someone higher up to actually admit that this is the reason.

Adam_swaine

06-03-2012, 20:47

Got bored, done a little looking into cmts, has to be upgrade work
http://ukinternetreport.co.uk/cmts/bexl.png
http://ukinternetreport.co.uk/cmts/details.php?area=bexl&town=Bexleyheath

http://ukinternetreport.co.uk/cmts/bmly.png
http://ukinternetreport.co.uk/cmts/details.php?area=bmly&town=Bromley

http://ukinternetreport.co.uk/cmts/brom.png
http://ukinternetreport.co.uk/cmts/details.php?area=brom&town=Bromsgrove

http://ukinternetreport.co.uk/cmts/dudl.png
http://ukinternetreport.co.uk/cmts/details.php?area=dudl&town=Dudley

http://ukinternetreport.co.uk/cmts/lewi.png
http://ukinternetreport.co.uk/cmts/details.php?area=lewi&town=Lewisham

http://ukinternetreport.co.uk/cmts/sand.png
http://ukinternetreport.co.uk/cmts/details.php?area=sand&town=Sandwell

http://ukinternetreport.co.uk/cmts/slam.png
http://ukinternetreport.co.uk/cmts/details.php?area=slam&town=South%20Lambeth

http://ukinternetreport.co.uk/cmts/telf.png
http://ukinternetreport.co.uk/cmts/details.php?area=telf&town=Telford

qasdfdsaq

06-03-2012, 20:50

Or major routing cockup (it's happened before)

Adam_swaine

06-03-2012, 21:00

Or major routing cockup (it's happened before)

How so, only ask as if it carries on I'll probably try and get it raised maybe email some higher up people, even though that probably won't help.

qasdfdsaq

06-03-2012, 21:05

Adam_swaine

06-03-2012, 21:15

Well if you look at all of those except Sandwell and Telford, each has at least one target in the same location that isn't seeing any difference in ping. The data is taking different routes, some of which are seriously bad ones.

Ok thanks for the explanation, is there a specific person I should be aiming this towards that can sort out the routing?

qasdfdsaq

06-03-2012, 21:21

No, not really. Unless you know them in person you're not going to get anywhere near the folk that do this kind of work. The executive office can give them a prod though.

Adam_swaine

06-03-2012, 23:07

No, not really. Unless you know them in person you're not going to get anywhere near the folk that do this kind of work. The executive office can give them a prod though.

Thanks, I've heard neil berkett actually does reply to emails. Or is there someone else who is better to contact.

---------- Post added at 23:07 ---------- Previous post was at 21:28 ----------

So tonight, same issue....started at 7, poor download speeds from there on, but it appears the download speed has come back a lot earlier tonight but still poor results from ping -n 100/pathping/tracert.

So not fully fixed yet but at least my dl speed is back.