Trace the Route Packets Take Between Two Hosts


traceroute

The traceroute command shows every step taken on the route from your machine to a specified host. Let's say you want to know why you can't get to www.granneman.com. You were able to load it just fine yesterday, but today's attempts to load the web page are timing out. Where's the problem?

$ traceroute www.granneman.com traceroute to granneman.com (216.23.180.5), 30 hops  max, 38 byte packets   1  192.168.0.1 (192.168.0.1)  1.245 ms  0.827 ms   0.839 ms   2  10.29.64.1 (10.29.64.1)  8.582 ms  19.930 ms    7.083 ms   3  24.217.2.165 (24.217.2.165)  10.152 ms  25.476   ms  36.617 ms   4  12.124.129.97 (12.124.129.97)  9.203 ms  8.003   ms  11.307 ms   5  12.122.82.241 (12.122.82.241)  52.901 ms  53.619  ms  51.215 ms   6  tbr2-p013501.sl9mo.ip.att.net (12.122.11.121)    51.625 ms  52.166 ms  50.156 ms   7  tbr2-cl21.la2ca.ip.att.net (12.122.10.14)    50.669 ms  54.049 ms  69.334 ms   8  gar1-p3100.lsnca.ip.att.net (12.123.199.229)    50.167 ms  48.703 ms  49.636 ms   9  * * *  10  border20.po2-bbnet2.lax.pnap.net  (216.52.255.101)  59.414 ms  62.148 ms  51.337 ms  11  intelenet-3.border20.lax.pnap.net  (216.52.253.234)  51.930 ms  53.054 ms  50.748 ms  12  v8.core2.irv.intelenet.net (216.23.160.66)   50.611 ms  51.947 ms  60.694 ms  13  * * *  14  * * *  15  * * * 


What do those * * * mean? Each one indicates a five-second timeout at that hop. Sometimes that could indicate that the machine simply doesn't understand how to cope with that traceroute packet due to a bug, but a consistent set of * indicates that there's a problem somewhere with the router to which v8.core2.irv.intelenet.net hands off packets. If the problem persists, you need to notify the administrator of v8.core2.irv.intelenet.net and let him know there's a problem. (Of course, it might not hurt to let the administrator of gar1-p3100.lsnca.ip.att.net know that his router is having a problem getting to border20.po2-bbnet2.lax.pnap.net as well, but it's not nearly the problem that intelenet.net is having.)

One other way to get around a problematic traceroute is to increase the number of hops that the command will try. By default, the maximum number of hops is 30, although you can change that with the -m option, as in traceroute -m 40 www.bbc.co.uk.

Tip

Actually, a better traceroute is mtr, which stands for Matt's traceroute. Think of it as a combination of ping and traceroute. If mTR is available for your Linux distribution, download and try it out. For more information, head over to www.bitwizard.nl/mtr.




Linux Phrasebook
Linux Phrasebook
ISBN: 0672328380
EAN: 2147483647
Year: 2007
Pages: 288

flylib.com © 2008-2017.
If you may any questions please contact us: flylib@qtcs.net