Subject:
|
Re: "jump.cgi" considered harmful ? (1)
|
Newsgroups:
|
lugnet.off-topic.geek
|
Date:
|
Fri, 1 Dec 2000 19:51:55 GMT
|
Viewed:
|
436 times
|
| |
|
|
In lugnet.off-topic.geek, Larry Pieniazek writes:
> I am not exactly sure how to interpret these (I ain't a network guy) but
> thought I'd post them:
Right, well, I am a network guy, so perhaps I can help to enlighten a bit...
> C:\WINNT\system32>ping www.bricksmiths.com
>
> Pinging bricksmiths.com [63.217.235.34] with 32 bytes of data:
>
> Reply from 63.217.235.34: bytes=32 time=100ms TTL=105
> Reply from 63.217.235.34: bytes=32 time=120ms TTL=105
> Request timed out.
> Reply from 63.217.235.34: bytes=32 time=101ms TTL=105
Well. 100-120ms seems like a lot to me, but then, we have no idea what kind of
connection you're sitting behind, and I've become accustomed to rather larger
pipes.
Still, what this does tell us is that you're dropping packets to
bricksmiths.com from whereever you're sitting, which I take from your
introduction to be a client's site in the UK, wholly unrelated to the host of
Bricksmiths or Lugnet.
> C:\WINNT\system32>ping www.lugnet.com
>
> Pinging lugnet.com [209.68.63.236] with 32 bytes of data:
>
> Reply from 209.68.63.236: bytes=32 time=100ms TTL=236
> Reply from 209.68.63.236: bytes=32 time=100ms TTL=236
> Reply from 209.68.63.236: bytes=32 time=100ms TTL=236
> Reply from 209.68.63.236: bytes=32 time=100ms TTL=236
Once again, 100ms seems high, but that's just my perception with no real clue
about the size of your pipe (huh huh). At least you dropped no packets.
> C:\WINNT\system32>tracert www.bricksmiths.com
>
> Tracing route to bricksmiths.com [63.217.235.34]
> over a maximum of 30 hops:
>
> 1 10 ms <10 ms <10 ms 172.17.0.2
> 2 <10 ms <10 ms 10 ms 170.10.10.10
> 3 10 ms 10 ms <10 ms intrtr1.tca.co.uk [193.133.105.80]
> 4 10 ms 10 ms 10 ms 193.128.72.238
> 5 10 ms 10 ms 10 ms gw14.lnd8.gbb.uk.uu.net [158.43.191.14]
> 6 10 ms 10 ms 10 ms ge5-0.cr1.lnd8.gbb.uk.uu.net [158.43.188.1]
> 7 10 ms 10 ms 10 ms pos0-0.cr1.lnd4.gbb.uk.uu.net [158.43.254.238]
> 8 10 ms 10 ms 10 ms ge2-0.cr2.lnd4.gbb.uk.uu.net [158.43.254.138]
> 9 10 ms 10 ms 10 ms pos0-2.cr2.lnd5.gbb.uk.uu.net [158.43.254.50]
> 10 10 ms 10 ms 10 ms so-7-0-0.XR2.LND2.alter.net [158.43.233.246]
> 11 10 ms 10 ms 10 ms SO-1-0-0.TR1.LND2.Alter.Net [146.188.7.230]
> 12 80 ms 90 ms 90 ms SO-6-0-0.IR1.DCA4.Alter.Net [146.188.8.169]
> 13 90 ms 90 ms 80 ms SO-0-0-0.IR1.DCA6.Alter.Net [146.188.13.33]
> 14 91 ms 90 ms 80 ms 118.at-4-1-0.TR1.DCA6.ALTER.NET • [152.63.10.122]
>
> 15 80 ms 90 ms 91 ms 287.at-5-0-0.XR1.TCO1.ALTER.NET [152.63.34.21]
> 16 90 ms 90 ms 90 ms 193.ATM11-0-0.BR1.TCO1.ALTER.NET • [146.188.160.73
> ]
> 17 90 ms 90 ms 90 ms mae-east.cais.com [192.41.177.85]
> 18 100 ms * 90 ms ser4-1.core2.mcl.cais.net [209.8.159.26]
> 19 90 ms 90 ms 90 ms pos1-1.core1.mcl.cais.net [63.216.0.17]
> 20 91 ms 90 ms 90 ms fe9-0-0.edge1.mcl.cais.net [209.8.159.81]
> 21 110 ms 100 ms 111 ms fe0-0.dsl2.mcl.cais.net [205.252.5.192]
> 22 * * * Request timed out.
> 23 130 ms 110 ms 141 ms 63.217.235.34
>
> Trace complete.
>
> (I did this one a few times, got very similar times in each position, always
> got a time out on hop 22)
Right, well, here we go. What this tells us is that the slowdown and dropped
packets to bricksmiths.com is occuring somewhere in the cais.net network.
However, I can't help but notice that you're going through uu.net/alter.net to
get there. In my experience, UUNet's backbone is a haven for congestion, and
that might well be what's causing the "dropped packets" via traceroute in the
cais network- but certainly not conclusively, and that's my own prejudice
showing through. Still, my experience with UUNet's network probably rivals
anyone on Lugnet's (unless there is a UUNet employee here).
At any rate, my diagnosis here is that cais.net sucks. Or UUNet sucks.
Something sucks, and it's not until you pass through UUNet's Tysons Corners
routers (what a big shock) that it sucks.
> C:\WINNT\system32>tracert www.lugnet.com
>
> Tracing route to lugnet.com [209.68.63.236]
> over a maximum of 30 hops:
>
> 1 <10 ms <10 ms <10 ms 172.17.0.2
> 2 <10 ms 10 ms <10 ms 170.10.10.10
> 3 <10 ms 10 ms <10 ms intrtr1.tca.co.uk [193.133.105.80]
> 4 10 ms 10 ms 10 ms 193.128.72.238
> 5 10 ms 10 ms 10 ms gw14.lnd8.gbb.uk.uu.net [158.43.191.14]
> 6 10 ms 10 ms 10 ms ge5-0.cr1.lnd8.gbb.uk.uu.net [158.43.188.1]
> 7 10 ms 10 ms 10 ms pos0-2.cr1.lnd4.gbb.uk.uu.net [158.43.254.246]
> 8 10 ms 10 ms 10 ms ge2-0.cr2.lnd4.gbb.uk.uu.net [158.43.254.138]
> 9 10 ms 10 ms 10 ms pos1-0.cr2.lnd5.gbb.uk.uu.net [158.43.254.249]
> 10 10 ms 10 ms 10 ms so-7-0-0.XR2.LND2.alter.net [158.43.233.246]
> 11 <10 ms <10 ms <10 ms SO-1-0-0.TR1.LND2.Alter.Net [146.188.7.230]
> 12 90 ms 91 ms 80 ms SO-6-0-0.IR1.DCA4.Alter.Net [146.188.8.169]
> 13 90 ms 90 ms 80 ms SO-0-0-0.IR1.DCA6.Alter.Net [146.188.13.33]
> 14 90 ms 90 ms 90 ms 118.at-4-1-0.TR1.DCA6.ALTER.NET • [152.63.10.122]
>
> 15 91 ms 90 ms 90 ms 287.at-5-0-0.XR1.TCO1.ALTER.NET [152.63.34.21]
> 16 90 ms 100 ms 91 ms 193.ATM6-0.GW3.PIT1.ALTER.NET [152.63.37.1]
> 17 100 ms 90 ms 100 ms pair-gw.customer.alter.net [157.130.48.162]
> 18 100 ms 90 ms 100 ms 192.168.1.4
> 19 100 ms 90 ms 100 ms lugnet.com [209.68.63.236]
>
> Trace complete.
And this, on the other hand, shows that you have no touble getting to Lugnet.
Since this trace also goes through UUNet's Tyson's Corners routers, and then
out to Pair's UUNet customer router in Pittsburgh, I would diagnose the above
dropped packets as meaning that cais.net is the sucky network. I assume,
though you don't mention it, that you had not trouble getting to Lugnet on
several tries, and this was around the same time (and I mean within minutes)
of the trouble reaching Bricksmiths. What would be a real test is doing
alternating traceroutes- one to Bricksmiths, one to Lugnet. If you keep
getting similar results, even when this is all done in the same time period, it
would be definitive that bricksmith.com's host's network sucks.
Of course, I have no idea what you think this proves about jump.cgi. Unless
you're on Lugnet's servers, you can't pin down how long it takes Lugnet to
reach bricksmiths.com... but from the above, I wouldn't be surprised if the
slowdown was being caused by cais.net.
eric
|
|
Message has 2 Replies:
 | | Re: "jump.cgi" considered harmful ? (1)
|
| Just as a follow up: I noticed that in Lar's traceroutes, the suck on cais.net happened after a Mae-East router. So, in the spirit of enlightened experimentation.... Here's a traceroute to lugnet.com from the Mae-East Looking Glass: Type escape (...) (24 years ago, 1-Dec-00, to lugnet.off-topic.geek)
| | | Re: "jump.cgi" considered harmful ? (1)
|
| (...) Am I misreading the big jump in time between 11 and 12 as transatlantic traffic? LND2 does sort of evoke "London" the same way that DCA evokes "Washington DC" (Tyson's Corners??) (...) Yes and yes. (...) I have no idea what it proves about (...) (24 years ago, 1-Dec-00, to lugnet.off-topic.geek)
|
Message is in Reply To:
| | Re: "jump.cgi" considered harmful ? (1)
|
| (...) I, at this client in the UK at a time when no one else is in the office and the connection SHOULD be lightly loaded (it's all new hardware here but we go somewhere else before it goes public) don't get quite the same results Mike does... same (...) (24 years ago, 1-Dec-00, to lugnet.off-topic.geek)
|
36 Messages in This Thread:
 
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
This Message and its Replies on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
