[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: basic ping diff with windows ping



On Tue, 26 Oct 2010 09:58:40 +0100, Adam Hardy wrote:

> Camaleón on 26/10/10 07:04, wrote:

>>> I didn't mean linux traceroute was quirky in execution - I just meant
>>> the options were not ideal for me. The dumbed-down version on windows
>>> was just right for my abilities and knowledge and what I wanted. But
>>> then if I hadn't used the windows traceroute first I might never have
>>> developed such preconceptions.
>> 
>> Well, Windows traceroute defaults to icmp while linux one seems to be
>> using udp which can be problematic with firewalls, so the windows
>> counterpart is a bit more "sensible" for today's routing diagnostics.
>> 
>> But true is that there is a slightly difference in the output we get
>> from a windows box traceroute and linux so besides the traceroute
>> utility itself there must be something in between which
>> interferes/alters the results.
> 
> I collated them in a spreadsheet just to check and I can't see any
> difference. I think if you were talking about the tracert output I
> showed before they were probably taken at different times completely.

(removing the traceroute output)

It's not about the time or old traceroutes, but the default options 
being used.

Let's take a closer look (shortened and deliberately non-breaking 
paragraphs).

Windows tracert uses icmp so we have to mimic that by telling linux 
traceroute to use the "-I" flag:

*****
Linux
*****

stt008:~# traceroute -I mktgw1.ibllc.com
traceroute to mktgw1.ibllc.com (208.245.107.9), 30 hops max, 40 byte 
packets
 1  192.168.0.69 (192.168.0.69)  1.255 ms  1.574 ms *
 2  * * *
 3  * * *
 4  * * *
 5  * * *
 6  * Xe5-1-0-0-grtmadno1.red.telefonica-wholesale.net (84.16.13.225)  45.896 ms Xe7-1-3-0-grtmadno1.red.telefonica-wholesale.net (84.16.13.213)  49.766 ms
...



(look at hop 2 -the router-, information on host/ip is complety missing, 
why? -timeout/iptables rule...?)




*******
Windows
*******

tracert mktgw1.ibll.com

Traza a la dirección mktgw1.ibll.com [208.73.210.29]
sobre un máximo de 30 saltos:

  1    <1 ms    <1 ms    <1 ms  192.168.0.69
  2    47 ms    48 ms    47 ms  10.6.28.1
  3    47 ms    48 ms    49 ms  130.Red-81-46-39.staticIP.rima-tde.net [81.46.39.130]
  4    49 ms    49 ms    46 ms  So0-0-0-0-grtmadde2.red.telefonica-wholesale.net [213.140.51.245]
  5    50 ms    49 ms    50 ms  p64-5-1-0.r21.mdrdsp01.es.bb.gin.ntt.net [129.250.9.109]
  6    74 ms    75 ms    75 ms  ae-3.r23.londen03.uk.bb.gin.ntt.net [129.250.6.53]
  7    93 ms   101 ms   101 ms  as-3.r22.amstnl02.nl.bb.gin.ntt.net [129.250.4.1.9]
...


Here I can see hop 2 without having to tweak nothing.

So in brief, to mimic the windows tracert behaviour we have to manually 
add a pause (-z):


stt008:~# traceroute -I -z 500 mktgw1.ibllc.com
traceroute to mktgw1.ibllc.com (208.245.107.9), 30 hops max, 40 byte 
packets
 1  192.168.0.69 (192.168.0.69)  1.016 ms  0.960 ms  0.953 ms
 2  10.6.28.1 (10.6.28.1)  47.184 ms  46.661 ms  47.888 ms
 3  81.Red-80-58-121.staticIP.rima-tde.net (80.58.121.81)  45.037 ms  47.192 ms  47.250 ms
 4  So-5-0-0-0-grtmadde2.red.telefonica-wholesale.net (84.16.9.165)  47.754 ms  45.032 ms  47.510 ms
 5  So-1-3-0-0-grtmadpe3.red.telefonica-wholesale.net (84.16.12.49)  47.978 ms  48.759 ms Xe7-1-3-0-grtmadad1.red.telefonica-wholesale.net (84.16.13.205)  48.325 ms
 6  Xe7-1-3-0-grtmadno1.red.telefonica-wholesale.net (84.16.13.213)  48.280 ms Xe5-1-0-0-grtmadno1.red.telefonica-wholesale.net (84.16.13.225)  48.900 ms Xe7-1-3-0-grtmadno1.red.telefonica-wholesale.net (84.16.13.213)  48.268 ms



Now hop 2 is identified. Question is "why" is that pause needed?

Greetings,

-- 
Camaleón


Reply to: