Slow page serve

retired_member16 · May 31, 2006

netserve said:
Whose SLA?

My employers.

netserve said:
How does 112ms ping time being better than most UK ISP's make any difference?

Was simply an observation. - nb - its now down to 60ms

netserve said:
And how can you get 13 hops on the same network? Are you saying that you're connected to the same ISP that delivers transit to Namedrive in CA and that there are no 3rd party networks in the middle?

Correct

netserve · Jun 1, 2006

firestars said:
Correct

How does that help anyone thats reported slow response times?

retired_member16 · Jun 1, 2006

netserve said:
How does that help anyone thats reported slow response times?

Because I can confirm that the server hotsting the sites is reporting correctly, 3 or 4 ISPs peering with us tested ok last night, whilst a few are bumbling around, handing off to 2 or 3 US ISPs and are suffering latency.

It is those UK based ISPs that have crappy peering points that are suffering latency, not NDs server.

netserve · Jun 1, 2006

firestars said:
whilst a few are bumbling around, handing off to 2 or 3 US ISPs and are suffering latency.

So you are seeing some odd latency then from your upstreams in CA?

firestars said:
It is those UK based ISPs that have crappy peering points that are suffering latency, not NDs server.

Hmm, peering at Linx in the UK delivers something like 80% of all worldwide routes directly and it's one of the largest peering points in the world. I don't think UK traffic that leaves the UK via Linx and then teleglobe.net can be considered crappy.

netserve · Jun 1, 2006

From Linx London...
6 hops to New York with Cable and Wireless.
Looks like at least 3 if not 4 different networks from New York to the colo facility for Name drive.

I've done several traces from different London points. There's several different tranatlantic carriers but in each case there's a fair mix of differen t transit providers on the US side of the connection.

I'm not suggesting its something that name drive are responsible for or can do anything about. Most UK traffic will head straight for NY and from their it's a lottery as to which networks the traffic gets carried over to arrive in CA.

1 g0-0.tr3.tcm.linx.net (195.66.248.126) 0 msec 0 msec 4 msec
2 fa0-0.transit1.thn.linx.net (195.66.248.243) 0 msec 4 msec 0 msec
3 insnet-transit.thn.linx.net (195.66.248.18) 0 msec 4 msec 0 msec
4 zcr2.lnt.cw.net (195.2.12.114) [AS 1273] 0 msec 0 msec 0 msec
5 so-1-2-0-dcr2.tsd.cw.net (195.2.10.130) [AS 1273] 4 msec 0 msec 4 msec
6 so-0-0-0-dcr1.nyk.cw.net (195.2.10.109) [AS 1273] 76 msec 76 msec 100 msec
7 GE3-0.PEERA-NYCMNY.IP.GROUPTELECOM.NET (198.32.160.78) 72 msec 76 msec 80 msec
8 POS7-0.WANB-MTRLPQ.IP.GROUPTELECOM.NET (66.59.191.177) [AS 6539] 80 msec 84 msec 84 msec
9 POS6-0.WANB-TOROON.IP.GROUPTELECOM.NET (216.18.63.197) [AS 6539] [MPLS: Label 12576 Exp 0] 92 msec 92 msec 96 msec
10 66.59.191.122 [AS 6539] 92 msec 92 msec 92 msec
11 h66-59-189-186.gtconnect.net (66.59.189.186) [AS 6539] 96 msec 96 msec 96 msec
12 ATM6-0-7.br0.win.mnsi.net (216.8.136.9) [AS 7057] 104 msec 108 msec
ATM6-0-6.br0.win.mnsi.net (216.8.136.17) [AS 7057] 108 msec
13 gw-nextdim.mnsi.net (216.8.136.142) [AS 7057] 112 msec 112 msec 112 msec
14 www.namedrive.com (216.8.177.27) [AS 13727] 112 msec 112 msec 116 msec

A-Wing · Jun 1, 2006

If it is any help to anyone here if the trace for me, I might have some latency at the moment due to a BRAS profile being created for the 8MBit regrade

1 192.168.1.1 (192.168.1.1) 3.194 ms 3.305 ms 2.986 ms
2 gauss-dsl.zen.net.uk (62.3.82.18) 22.137 ms 22.015 ms 21.643 ms
3 deleuze-ge-0-1-0-1.hq.zen.net.uk (62.3.80.161) 21.941 ms 22.155 ms 24.952 ms
4 lotze-ge-0-0-0-0.hq.zen.net.uk (62.3.80.57) 21.538 ms 25.700 ms 35.726 ms
5 zermelo-so-0-0-0-0.sp.zen.net.uk (62.3.80.82) 25.951 ms 24.070 ms 24.668 ms
6 zubiri-so-0-0-0-0.ny.zen.net.uk (62.3.80.86) 104.837 ms 106.020 ms 104.922 ms
7 GE3-0.PEERA-NYCMNY.IP.GROUPTELECOM.NET (198.32.160.78) 104.062 ms 105.433 ms 104.839 ms
8 POS7-0.WANB-MTRLPQ.IP.GROUPTELECOM.NET (66.59.191.177) 113.921 ms 112.615 ms 112.964 ms
9 POS6-0.WANB-TOROON.IP.GROUPTELECOM.NET (216.18.63.197) 128.700 ms 127.346 ms 130.802 ms
10 216.18.63.14 (216.18.63.14) 122.475 ms 122.467 ms 121.158 ms
11 h66-59-189-186.gtconnect.net (66.59.189.186) 124.913 ms 124.835 ms 125.262 ms
12 ATM6-0-6.br0.win.mnsi.net (216.8.136.17) 128.822 ms ATM6-0-7.br0.win.mnsi.net (216.8.136.9) 128.941 ms 131.237 ms
13 gw-nextdim.mnsi.net (216.8.136.142) 130.972 ms 128.524 ms 130.016 ms
14 ptr-216-8-177-27.ptr.nextdimensioninc.com (216.8.177.27) 134.141 ms 131.092 ms 133.451 ms

retired_member16 · Jun 2, 2006

netserve said:
So you are seeing some odd latency then from your upstreams in CA?

Because im able to log onto route servers of a few 00s of ISPs in the UK and trace from there.

netserve said:
Hmm, peering at Linx in the UK delivers something like 80% of all worldwide routes directly and it's one of the largest peering points in the world. I don't think UK traffic that leaves the UK via Linx and then teleglobe.net can be considered crappy.

I never said anything about Linx, and anyhow, you only get your transatlantic routing from there, once you hit the US, you may need to jump across a few peering points before you get to your destination. Unless you go directly and not through Telehouse of course.