Network problems Tonight ? - Page 2

Ignition · 07-08-2005, 08:40

Quote:

Originally Posted by IanUK

The first large hop (212.250.14.62) appears to belong to NTL (according to Neotrace).
Looks like a problem they need to deal with.

inetnum: 212.250.14.0 - 212.250.15.255
netname: NTL
descr: NTL Infrastructure - Private Peering

Looks like the direct private link between ntl and Cogentco was saturated or Cogentco were having problems, although it's using ntl IP addresses that IP address is actually the Cogentco side of the link.

Unlikely to be saturation though that would have caused some other entertainment as well.

IanUK · 07-08-2005, 10:32

I'd guess it is saturation, the day is moving along and the congestion is starting again:

Target Name: www.filefront.com
IP: 66.117.8.41
Date/Time: 07/08/2005 11:32:03
2 5 ms [10.82.176.1]
3 8 ms wapk-t2cam1-b-ge-wan41-112.inet.ntl.com [80.1.170.185]
4 7 ms popl-t2core-b-ge-wan64.inet.ntl.com [62.255.81.193]
5 16 ms pop-bb-b-so-700-0.inet.ntl.com [213.105.174.117]
6 10 ms tele-ic-1-so-010-0.inet.ntl.com [62.253.185.82]
7 47 ms [212.250.14.62]
8 119 ms p10-0.core01.bos01.atlas.cogentco.com [130.117.0.46]
9 145 ms p5-0.core01.ord01.atlas.cogentco.com [66.28.4.110]
10 140 ms p15-0.core02.ord01.atlas.cogentco.com [66.28.4.62]
11 151 ms p12-0.core01.mci01.atlas.cogentco.com [66.28.4.33]
12 163 ms p12-0.core01.iah01.atlas.cogentco.com [66.28.4.105]
13 195 ms p14-0.core01.san01.atlas.cogentco.com [66.28.4.6]
14 195 ms p4-0.core01.lax01.atlas.cogentco.com [66.28.4.77]
15 197 ms filefront-allyance.dmarc.cogentco.com [38.112.29.134]
16 199 ms node.www.la.ioerror.com [66.117.8.41]

Up to 47 on that hop now.

Ignition · 07-08-2005, 12:02

OK Have found fault, can't promise it'll be sorted today but things should seem better tomorrow

IanUK · 07-08-2005, 12:04

Thanks, your help is much appreciated.

IanUK · 07-08-2005, 14:10

I know you are already looking into this,but I thought this might help - I've had a response to my trouble ticket with EV1 - they seem to think its the Poplar Proxy.

Here is an extract from what they sent me:

---------
Testing to that proxy address was terrible.

telnet@240.125#ping 62.255.64.12 co 1000 size 1420 source 66.98.246.1
Sending 1000, 1420-byte ICMP Echo to 62.255.64.12, timeout 5000 msec, TTL 64
Success rate is 97 percent (974/1000), round-trip min/avg/max=180/203/220 ms.

(very bad packet loss - 3%)

The trace to that host:

240.125#trace 62.255.64.12
Type Control-c to abort
Tracing the route to IP node 62.255.64.12 from 1 to 30 hops

1 <1 ms <1 ms <1 ms gphou-66-98-241-4.ev1.net [66.98.241.4]
2 <1 ms <1 ms <1 ms gphou-66-98-240-10.ev1.net [66.98.240.10]
3 <1 ms <1 ms <1 ms g0-3.na21.b015619-0.iah01.atlas.cogentco.com [38.112.25.21]
4 1 ms 1 ms 1 ms g10-3-0.core01.iah01.atlas.cogentco.com [66.28.64.65]
5 15 ms 15 ms 15 ms p13-0.core01.mci01.atlas.cogentco.com [66.28.4.106]
6 26 ms 26 ms 29 ms p5-0.core02.ord01.atlas.cogentco.com [66.28.4.34]
7 39 ms 26 ms 26 ms p15-0.core01.ord01.atlas.cogentco.com [66.28.4.61]
8 58 ms 56 ms 47 ms p14-0.core01.bos01.atlas.cogentco.com [66.28.4.109]
9 118 ms 117 ms 118 ms p3-0.core01.lon02.atlas.cogentco.com [130.117.0.45]
10 212 ms 209 ms 209 ms tele-ic-1-ge-001-217.inet.ntl.com [212.250.14.61]
11 211 ms 212 ms 218 ms pop-bb-b-so-430-0.inet.ntl.com [62.253.185.81]
12 210 ms 209 ms 214 ms popl-t2core-b-pos81.inet.ntl.com [213.105.174.118]
13 208 ms 203 ms 213 ms popl-cache-9.server.ntli.net [62.255.64.12]

The loss seems to be between hops 10 and 11 outbound. I would need an inbound trace from the proxy to the server to be sure.

good to hop 10:

240.125#ping 212.250.14.61 size 1420 co 1000
Sending 1000, 1420-byte ICMP Echo to 212.250.14.61, timeout 5000 msec, TTL 64
Success rate is 100 percent (1000/1000), round-trip min/avg/max=109/109/128 ms.

(1,000 pings in both directions, perfect, 109 ms avg latency)

bad to hop 11:

240.125#ping 62.253.185.81 size 1420 co 100
Sending 100, 1420-byte ICMP Echo to 62.253.185.81, timeout 5000 msec, TTL 64
Success rate is 98 percent (98/100), round-trip min/avg/max=204/214/250 ms.

(loss and higher latency and high jitter)
----------

I have to say that I'm using Swansea proxies again as the Poplar one is awful again currently

Thanks again for your help.

cheers

Ian

Ignition · 07-08-2005, 14:31

From Poplar core (the same that they show in their trace)

To: 62.253.185.81

Using pings of 1420 bytes as above:

Sending 1000, 1420-byte ICMP Echos to 62.253.185.81, timeout is 2 seconds:
Success rate is 100 percent (1000/1000), round-trip min/avg/max = 1/2/36 ms

To the hop 10 in the trace above:

Sending 1000, 1420-byte ICMP Echos to 212.250.14.61, timeout is 2 seconds:
Success rate is 100 percent (1000/1000), round-trip min/avg/max = 1/2/40 ms

To Cogent 1st hop, the peering router which tele-ic-1 connects to, in their trace this is 9 118 ms 117 ms 118 ms p3-0.core01.lon02.atlas.cogentco.com [130.117.0.45]

Sending 1000, 1420-byte ICMP Echos to 130.117.0.45, timeout is 2 seconds:
Success rate is 94 percent (940/1000), round-trip min/avg/max = 20/74/300 ms

Not much else to say really, they are wrong, worked ok at one point for them when they were testing however their traceroute shows where the fault lies:

9 118 ms 117 ms 118 ms p3-0.core01.lon02.atlas.cogentco.com [130.117.0.45]
10 212 ms 209 ms 209 ms tele-ic-1-ge-001-217.inet.ntl.com [212.250.14.61]

Ping response shown by them is normal for Netapps.

IanUK · 07-08-2005, 15:48

LOL, I'll let you techies battle it out between yourselves, most of that goes waaay over my head

__________________

That hop suddenly looks better.

Target Name: www.filefront.com
IP: 66.117.8.41
Date/Time: 07/08/2005 16:32:57

2 6 ms 6 ms [10.82.176.1]
3 7 ms 7 ms wapk-t2cam1-b-ge-wan41-112.inet.ntl.com [80.1.170.185]
4 7 ms 8 ms popl-t2core-b-ge-wan64.inet.ntl.com [62.255.81.193]
5 8 ms 8 ms pop-bb-b-so-700-0.inet.ntl.com [213.105.174.117]
6 8 ms 10 ms tele-ic-1-so-010-0.inet.ntl.com [62.253.185.82]
7 16 ms 15 ms [212.250.14.62]
8 84 ms 88 ms p10-0.core01.bos01.atlas.cogentco.com [130.117.0.46]
9 105 ms 107 ms p5-0.core01.ord01.atlas.cogentco.com [66.28.4.110]
10 107 ms 106 ms p15-0.core02.ord01.atlas.cogentco.com [66.28.4.62]
11 118 ms 118 ms p12-0.core01.mci01.atlas.cogentco.com [66.28.4.33]
12 133 ms 131 ms p12-0.core01.iah01.atlas.cogentco.com [66.28.4.105]
13 165 ms 163 ms p14-0.core01.san01.atlas.cogentco.com [66.28.4.6]
14 164 ms 165 ms p4-0.core01.lax01.atlas.cogentco.com [66.28.4.77]
15 213 ms 217 ms filefront-allyance.dmarc.cogentco.com [38.112.29.134]
16 215 ms 214 ms node.www.la.ioerror.com [66.117.8.41]

Downloads are still slow through NTL though (sub 20k)
Downloading through AOL gives full speed (375k)

Maybe this isn't a fix, just a blip ?
__________________

File Test

This file is on my server with EV1.

Target Name: www.froopter.com
IP: 66.98.246.48
Date/Time: 07/08/2005 16:50:10
2 7 ms 8 ms [10.82.176.1]
3 8 ms 8 ms wapk-t2cam1-a-ge-wan41-112.inet.ntl.com [80.1.170.57]
4 8 ms 8 ms popl-t2core-a-ge-wan64.inet.ntl.com [62.255.81.65]
5 9 ms 7 ms pop-bb-a-so-330-0.inet.ntl.com [213.105.174.113]
6 7 ms 12 ms pop-bb-b-ae0-0.inet.ntl.com [213.105.172.77]
7 8 ms 9 ms tele-ic-1-so-010-0.inet.ntl.com [62.253.185.82]
8 16 ms 25 ms [212.250.14.62]
9 85 ms 87 ms p10-0.core01.bos01.atlas.cogentco.com [130.117.0.46]
10 104 ms 108 ms p5-0.core01.ord01.atlas.cogentco.com [66.28.4.110]
11 105 ms 114 ms p15-0.core02.ord01.atlas.cogentco.com [66.28.4.62]
12 119 ms 118 ms p12-0.core01.mci01.atlas.cogentco.com [66.28.4.33]
13 130 ms 134 ms p12-0.core01.iah01.atlas.cogentco.com [66.28.4.105]
14 130 ms 131 ms g0-2.na21.b015619-0.iah01.atlas.cogentco.com [66.28.64.66]
15 132 ms 133 ms everyones_internet_2.demarc.cogentco.com [38.112.25.158]
16 117 ms 116 ms gphou-66-98-241-27.ev1.net [66.98.241.27]
17 114 ms 115 ms gphou-66-98-241-125.ev1.net [66.98.241.125]
18 117 ms 117 ms srv1.froopter.com [66.98.246.48]

I get about 80k max dropping to 20k most of the time, from a variety of NTL proxies, but get full speed (over 350k) from non NTL Proxies and an AOL connection - the poor hop appears to have been fixed but my speeds are still rubbish using NTL. I can now get full speed from collecting my mail from the server, but not from the web - which now makes me think its the proxies

FTP'ing also gives me full speed, I'm sorry but I tend to support EV1 here and say its the proxies that are at fault, direct connections seem to work fine - even with my little techy knowledge.

If anyone on a 3 meg connection has the time to test the speed by downloading this file using a web browser from my server thru NTL it would be much appreciated.

http://www.froopter.com/radeondriver.test

Its a renamed 10 meg exe file (a Radeon Driver) and completely harmless - I'm not asking you to run it, just download it then delete it, and see what speeds you get.

Many Thanks

Ian

07-08-2005, 16:13

Quote:

Originally Posted by IanUK

http://www.froopter.com/radeondriver.test

Its a renamed 10 meg exe file (a Radeon Driver) and completely harmless - I'm not asking you to run it, just download it then delete it, and see what speeds you get.

Many Thanks

Ian

Getting 365 kB here in Warrington

total

10.6meg in 27 sec

IanUK · 07-08-2005, 16:19

Thanks, can I ask which proxy you are using (so I can try it from here)
or PM me please if you don't want to publish it.

Thanks

Ian

07-08-2005, 16:22

Quote:

Originally Posted by IanUK

Thanks, can I ask which proxy you are using (so I can try it from here)
or PM me please if you don't want to publish it.

Thanks

Ian

Its easy to work out. I am in Warrington so i am on bagu-cache-4.server.ntli.net

hope that helps

IanUK · 07-08-2005, 16:24

Thanks, guess what - I get 370k with that Proxy too !!
I'd been getting slower speeds from lots of sites and put it down to the sites being busy, but that proxy gives me full speed everywhere again !

Looks like Poplar and Colchester and Swansea (the others I tried) are not giving full speeds again

I do wish instead of speed increases NTL would fix the proxies, I can't see me getting any benefit from 10mb when I can't even fill 3 meg without continual proxy swapping

07-08-2005, 17:11

Quote:

Originally Posted by IanUK

Thanks, guess what - I get 370k with that Proxy too !!
I'd been getting slower speeds from lots of sites and put it down to the sites being busy, but that proxy gives me full speed everywhere again !

Looks like Poplar and Colchester and Swansea (the others I tried) are not giving full speeds again

I do wish instead of speed increases NTL would fix the proxies, I can't see me getting any benefit from 10mb when I can't even fill 3 meg without continual proxy swapping

Glad that helps you

IanUK · 07-08-2005, 17:15

Thanks again, that bad hop is building up again, so looks like no fix yet (that would have been quick!)

I'll shut up now until they have a chance to fix it !

jtwn · 07-08-2005, 17:16

How are you getting 3mbit on an AOL connection? Have I missed something

IanUK · 07-08-2005, 17:24

I'm not

I have a 3meg NTL connection, and use the AOL client in 'Bring your own access' mode on the top. Using that client routes all content through AOL's proxies rather than NTL's, all sites work at full speed through it, which is much better than the NTL Poplar Proxies which are quite frankly awful.