Traceroute strangeness

[sticky_chicken]

For a while now, I've noticed that I'm getting occasional oddities with traceroute from windoze.

The symptoms I'm getting are that the traceroute dies within swiftel or one of their providers - an example is below:

E:\WINDOWS\system32>tracert eon.net.au

Tracing route to eon.net.au [203.55.198.1]
over a maximum of 30 hops:

1 <1 ms <1 ms <1 ms eth1.e-mailme.org [10.0.0.1]
2 19 ms 19 ms 19 ms p-nya.swiftel.com.au [202.154.95.173]
3 19 ms 19 ms 20 ms p062.swiftel.com.au [202.154.83.62]
4 * * * Request timed out.
5 * * * Request timed out.
6 * * * Request timed out.
7 * * * Request timed out.
8 * * * Request timed out.

whilst a traceroute to the same host from my linux gateway works - see below:

sporran[~]trace 216.118.89.6
traceroute to 216.118.89.6 (216.118.89.6), 30 hops max, 38 byte packets
1 p-nya.swiftel.com.au (202.154.95.173) 29.272 ms 16.825 ms 16.425 ms
2 p062.swiftel.com.au (202.154.83.62) 17.169 ms 18.231 ms 16.447 ms
3 ge-2-0-0.401.ar2.syd1.gblx.net (203.192.141.121) 21.609 ms 16.466 ms 16.439 ms
4 ge0-0-0-1000M.cr2.SYD1.gblx.net (203.192.136.85) 16.703 ms 15.918 ms 16.437 ms
5 pos3-3-155M.cr2.SNA1.gblx.net (203.192.136.114) 189.653 ms pos2-1-155M.cr1.LAX1.gblx.net (203.192.136.106) 179.618 ms 176.173 ms
6 pos1-0-2488M.cr2.JFK1.gblx.net (67.17.72.10) 269.215 ms 265.127 ms 256.563 ms
7 so3-0-0-2488M.ar1.JFK1.gblx.net (67.17.72.18) 254.949 ms 254.850 ms 254.174 ms
8 Net-Access-Corporation-Net-Access-Corp.so-0-2-1.ar1.JFK1.gblx.ne (64.211.83.62) 362.537 ms 343.651 ms 343.656 ms
9 0.so-0-3-0.gbr1.oct.nac.net (209.123.11.233) 353.666 ms 360.966 ms 358.561 ms
10 95.gi-1-1.msfc1.oct.nac.net (64.21.102.2) 349.445 ms 357.356 ms 341.957 ms
11 216.118.89.6 (216.118.89.6) 358.656 ms 353.591 ms 357.669 ms


Is anyone else who's running windoze behind a linux gateway seeing this?



:confused:

[djc]

Windows uses 92 byte length ICMP packets for trace routes. Swiftel and/or their upstream providers likely block these packets because of the recent viruses. Linux uses a different method or a different packet size to do the traceroutes.

[sticky_chicken]

Windows uses 92 byte length ICMP packets for trace routes. Swiftel and/or their upstream providers likely block these packets because of the recent viruses. Linux uses a different method or a different packet size to do the traceroutes.

Cool - thanks for your answer. I suspected as much.

[Bradley Amm]

eon use swiftel as their bandwidth provider

[22chris]

Always use linux for doing your traceroutes they'll always come out more accurate and are better to use in the long run.

[Bradley Amm]

no linux is gay. too hard to work out

its for geeks

[sticky_chicken]

Originally posted by Bradley Amm
no linux is gay. too hard to work out

its for geeks

Well, I'm glad to see that you've got an open mind about this.

If everyone else thought the way you did, then we'd still believe

1) That the earth is flat

2) If humans were meant to fly that we'd have wings

3) Putting a man on the moon is impossible.


Things in life that are worth accomplising, always seem hard at first.

If it seems too hard, then I suggest you go back to the sandpit and play in that.

[wct296]

Well there is an alternative.. Download the attached zip file ( or goto http://www.dslreports.com/forum/remark,8563796~root=speakeasy~mode=flat)

It has an alternative tracert file for windows that doesnt have the same error.

It works well ... Very handy...