20-11-2009 16:42 - edited 20-11-2009 16:43
Dont know if this will give anyone who has a good technical knowledge of pings and trace routes any ideas ?, but i was thinking why dont we start pinging the virgin servers and do visual traceroute tests, maybe it would shed some light on where in the virgin systems we are getting slowed down if in fact it is the case, now ive only just started tinkering with it and ive found that pinging my imediate server.
I decided to (and i will do many more destinations) do a random traceroute a usa server to see if there is a slow down before it leaves the virgin networks and i did notice some blips where the ms jumped during the hops through the various virgin servers, maybe others could run some tests and post a few results, because if its going through a particular hub where the pings spike it may show virgin where to take a look, what do others think ? is it worth looking into ?
first i goto http://visualroute.visualware.co.uk/ and run a trace route and run cmd and ping whatever server with -t on the end for continuous pings and do this for each stage in the route to see where the spikes occur.
so 1st hop is me to base server @ basildon on 18.104.22.168 i get <1ms but i had about 6 times in a 100 jump up to 87ms
next which is still in the basildon hub on 22.214.171.124 i get between 4-8ms with spikes upto 79ms every 10-20 seconds. so i now know in on the basl-cmts-09 server (Thanks to ignition on this point)
next (still inside the basildon hub basi-pc200 on 126.96.36.199) i get 4-8 ms but get spikes every 10-20 seconds upto 238ms but most spikes average 50-70ms.
next 188.8.131.52 which is now on the manchester hub i get 6-8ms but loads more spikes every 2-3 seconds where it mostly jumps 30-75ms. (this ip being the jumpiest so far)
still in the manchester hub on 184.108.40.206 the pings are 14-15ms but again with much more spikes than the basildon hub.
now intrestingly it hops over to washington usa on 220.127.116.11 and im getting mostly 14-15 ms again but i did two ping tests side by side manchester and usa and when manchester hub spiked it had a knock on effect to the usa but the usa spicked slightly lower ms.
so should some of us gather a bit of data and see if a pattern emerges, cos if there is a pattern we could at least get VM to check it out.
20-11-2009 19:06 - edited 20-11-2009 19:11
Are you using wireless?
18.104.22.168 is your router / PC. Spikes to that point to a problem with yourown kit
Also, Visualroute's guesses of where things are is wrong, and there's nothing wrong with seeing spikes to routers on the network, apart from your own
20-11-2009 20:07 - edited 20-11-2009 20:27
nope im not on wireless, everything is hard wired.
not my kit, as virgin say its their kit so their problem.
maybe the kit they supply is not up to the job then.
but the spikes are showning serverside mostly, and a lot of it from the manchester hub.
I only get a couple of spikes on 22.214.171.124, but it rises expedentially as it travels through their networks.
It wouldnt suprise me if its the modems and routers not upto the job though, ive had both replaced but when techs visit they dont run any thorough tests apart from check power levels, plus the modems and the routers are the cheaper end of the scale.
(dont forget they upgraded me to 50Mb/s while i was still on my surfboard 5100 modem & the speeds worked perfectly until they installed the black modems)
They are currently keeping checks on my line to see if its a "server fault or a configuration fault" they said looking at my monthly speeds on the surface the speeds look ok but looking day by day the speeds are down, so that just makes no sense cos the day by day makes up the monthly average speeds.
Im more bothered about the lack of stability over speeds, i want a line the stays on not one that goes offline several times a day.
im also on the sam-knows project so it records the data so i can back up what i say to them.
on 20-11-2009 20:33
The routers on the core network are great and very capable. Perhaps it'd be worth your while using something like the reverse traceroute tool here.
Should show things a bit more clearly. I would expect to see spikes on the routers themselves as answering your pings / traceroutes is both low priority to them and rate limited.
That TBB tool might give you more value and pinpoint issues
20-11-2009 21:47 - edited 20-11-2009 21:51
Which Washington was it? You may not know this but there is a Washington in the UK (Americans stole the name like the language). Same goes for most of their citys/states actually.
The 200 IP makes me think its a USA IP but I'm no expert on those so I'm just asking if you know for sure.
on 20-11-2009 22:37
Some routers will occasionally show poor response times. This is because if a router is busy doing things such as routing traffic, or updating its routing data, it will put preference to this rather than respond to non-critical network tests.
You are about to request a traceroute that may be interpreted as an 'attack' from www.thinkbroadband.com by a firewall on your computer.
# Host Sent Recvd Best Avg Worst
1 thinkbroadband-gw2.core-rs3.thdo.ncuk.net 15 15 0ms 0ms 1ms
2 gi0-2-10-star1.bdr-rt2.thdo.ncuk.net 15 15 0ms 0ms 0ms
3 linx-gw2.router.ntli.net 15 15 0ms 1ms 6ms
4 bre-bb-b-as0-0.network.virginmedia.net 15 15 1ms 1ms 5ms
5 pop-bb-a-as7-0.network.virginmedia.net 15 15 1ms 18ms 123ms
6 lutn-t3core-1a-as0-0.network.virginmedia.net 15 15 2ms 2ms 3ms
7 heme-cmts-09-gigaether-150.network.virginmedia.net 15 15 3ms 4ms 7ms
There you go,no USA traces but I ain`t complaining with pings like that,and no drops for the last 5 days,and only 1 T3 error in my modem log file.
So I quess it ain`t the VM servers.
on 21-11-2009 09:59
How does hop 8 look though, that's the most likely cause of failure between hop 7 and hop 8. Censor as required, naturally!
on 21-11-2009 11:59
Well wherever the problem lays i hope they manage to get it sorted soon cos it is ridiculus to have so many offline and speed issues from the day the 50mb modems were installed. ive had one good week of top speeds and no offline times and thats as i said before on the old motorola surfboard5100 modem when they sent the signal through to my modem for the upgrade a week before the tech arrived with the new modem.
22-11-2009 04:57 - edited 22-11-2009 05:06
Tracing route to www.bbc.net.uk [126.96.36.199]
over a maximum of 30 hops:
1 <1 ms <1 ms <1 ms 192.168.*.*
2 5 ms 7 ms 7 ms heme-cmts-09-lback-20.network.virginmedia.net [8
3 7 ms 7 ms 9 ms lutn-t3core-1a-ge-300-2545.network.virginmedia.n
4 7 ms 9 ms 6 ms pop-bb-a-as5-0.network.virginmedia.net [213.105.
5 12 ms 11 ms 7 ms nth-bb-b-as3-0.network.virginmedia.net [213.105.
6 13 ms 11 ms 13 ms tele-ic-1-as0-0.network.virginmedia.net [62.253.
7 10 ms 13 ms 11 ms pos6-1.rt0.thdo.bbc.co.uk [188.8.131.52]
8 13 ms 14 ms 15 ms 184.108.40.206
9 13 ms 11 ms 11 ms 220.127.116.11
10 12 ms 11 ms 11 ms www-nonuk-vip.telhc.bbc.co.uk [18.104.22.168]
Adding them up the whole distance comes to 100ms approx.
I use OpenDNS,gigabit ethernet adapters,cat 5e,and use a Belkin N1 Vision router with gigabit WAN and 4 LAN ports also gigabit.
It has an LCD display that gives a lot of info screens and a clock with day & date.