Tracing...

[hometown]

I long ago was taught by my loved one about how to trace route...:)

I trace some web pages for fun but here comes the error:
Unable to resolve target system name
What s the problem with that ? I couldnot know where those pages come from ? Are there any other ways to get the exact location of those pages this or how to get information about those ?

Thanks a lot...:)
Regards,

[khp]

You can't trace web pages (it simply dosen't make sense).
You can trace, at least part of the route, to the web server that hosts a page.

for example you can trace www.codeguru.com, using the command tracert www.codeguru.com

Asuming that you are using windows.

[hometown]

Originally posted by khp
You can't trace web pages (it simply dosen't make sense).
You can trace, at least part of the route, to the web server that hosts a page.

for example you can trace www.codeguru.com, using the command tracert www.codeguru.com

Asuming that you are using windows.

Thanks a lot for your answer, Khp...:)

But what does "*" mean when tracing and why system says "request time out" ?

Thanks...:)
-Nina

[khp]

It means that the next network node on the path to the server is not responding to the trace request.

[hometown]

Originally posted by khp
It means that the next network node on the path to the server is not responding to the trace request.
So, do you know how to fix it ?

[khp]

Sure change the router/server configuration so it responds to trace requests.

Of course this assumes that you have full control over that particular router/server, which I would suspect that you don't have.

[hometown]

Originally posted by khp
Sure change the router/server configuration so it responds to trace requests.

Of course this assumes that you have full control over that particular router/server, which I would suspect that you don't have.
Thanks....
But I have another question....Do you know what can cause such a problem ? I mean the server's problem that disables my access to a certain page...
Thanks again...

[khp]

Originally posted by hometown
Do you know what can cause such a problem ?

Either the server/router has crashed, someone has cut the cable, or the router/server is bloking trace requests. Assuming that you are able to get through with normal http requests, it's looks like it's the latter case.

Originally posted by hometown
I mean the server's problem that disables my access to a certain page...
Thanks again...

Not necessarily, traceing the path to a server and making a http request for a cirtain page on a server are completly seperate issues.

Of course if the server doesn't respond to http requests and doesn't respond to ping, it seems likely that the server has gone offline, or someone has cut the network connection between you and the server.

[hometown]

....

[hometown]

I have some questions...Please help me...
When I trace the route, there are also alot of different unknown IPs and their host names showing up, could you tell me why that happened ? Why can I invoke those kids before I invoke their parent *?

Thanks alot for any help...
-Home


* So sorry, i dont know what terms I should use in place of them, I am not in the field...:)

[khp]

Originally posted by hometown
When I trace the route, there are also alot of different unknown IPs and their host names showing up,


What do you mean by unknow IPs and hostname. What more do you expect to know about a system ?

Originally posted by hometown
Why can I invoke those kids before I invoke their parent *?


I don't quite understand what you mean by invoke, kids and parents. There are no kids or parents, just nodes on a path between you and the target. And you don't invoke them, you query them, for their name and IP.

Originally posted by hometown
* So sorry, i dont know what terms I should use in place of them, I am not in the field...:)

Then use some examples to ******** what you mean.

[hometown]

Originally posted by khp
What do you mean by unknow IPs and hostname. What more do you expect to know about a system ?
I don't quite understand what you mean by invoke, kids and parents. There are no kids or parents, just nodes on a path between you and the target. And you don't invoke them, you query them, for their name and IP.

I called like that because I dont know how to call them, now I know they are nodes and I querry their names and IPs...:)
We can know only about their names and IPs only ? are there any other things we can know about them ? I am sorry I dont know what examples I have to use now....
But as what you told me...for example:
Me->Node_01->Node_02->Node_03->....->Node_0n-> Target
like a graph right ? But is there any path that is just Me->Target, no node in between ????? Oh, for example I have 2 computer and I connect them together, both will create such a path, right ?
By the way, the red graph is directed graph??? If so, how can I compute the length between Me and Target or my question should be Is this computation possible before asking so?

Thanks Khp alot for your help....:)

Regards,
-Nina

[khp]

Originally posted by hometown
We can know only about their names and IPs only ? are there any other things we can know about them ?

What more would you want to know ?. Somewhere one the web (I don't remember where) there are loopup services that lets you lookup the owner of a domain name or IP number (probably some ICANN or Verisign page)

Originally posted by hometown
I am sorry I dont know what examples I have to use now....
How can you not have an example ?, what caused you to ask in the first place ?.

Originally posted by hometown
But as what you told me...for example:
Me->Node_01->Node_02->Node_03->....->Node_0n-> Target
like a graph right ?

It's more like a path on a graph, the internet is an interconnected web of millions of nodes, the path you see is just one possible path through that graph.

Originally posted by hometown
But is there any path that is just Me->Target, no node in between ?????

Yes, as long as you and the target is directly connected (nothing but switches and hubs between you). But on the internet, you will always have to pass through a router belonging to your isp, before you can go anywhere else.

Originally posted by hometown
Oh, for example I have 2 computer and I connect them together, both will create such a path, right ? Yes.

Originally posted by hometown
By the way, the red graph is directed graph?

In theory, I suppose it is, I think most connecttions work in duplex, these days. The only counter example I can think of, is with people using sattelite modems. But of course there is no need for 2 way communications to follow the same route in both directions. When I send a trace request to a node, between me and some target, it's response may come to me by a completly different route, than the route I'am seeing.

Originally posted by hometown
my question should be Is this computation possible before asking so?


No you can't know ahead of time, since you never have complete knowledge of the network graph that makes up the internet (and noone else does). But of course there are more ways to calculate the length (hop count) between two points (But you can't even be sure it will follow the same route every time). You could play around with the hop count on a ping request (see how sort you can make it, and still get a response)

[hometown]

Originally posted by khp
What more would you want to know ?. Somewhere one the web (I don't remember where) there are loopup services that lets you lookup the owner of a domain name or IP number (probably some ICANN or Verisign page)
How can you not have an example ?, what caused you to ask in the first place ?.
I am sorry I didnot think of that address at that time...
Here it is
http://neic.usgs.gov/neis/FM/neic_zebr_q.html
It's more like a path on a graph, the internet is an interconnected web of millions of nodes, the path you see is just one possible path through that graph.
Yes, as long as you and the target is directly connected (nothing but switches and hubs between you). But on the internet, you will always have to pass through a router belonging to your isp, before you can go anywhere else.

Thanks a lot...
In theory, I suppose it is, I think most connecttions work in duplex, these days. The only counter example I can think of, is with people using sattelite modems. But of course there is no need for 2 way communications to follow the same route in both directions. When I send a trace request to a node, between me and some target, it's response may come to me by a completly different route, than the route I'am seeing.
No you can't know ahead of time, since you never have complete knowledge of the network graph that makes up the internet (and noone else does). But of course there are more ways to calculate the length (hop count) between two points (But you can't even be sure it will follow the same route every time). You could play around with the hop count on a ping request (see how sort you can make it, and still get a response)
Could you tell me why it doesnot follow the same route ? Why does that happen ? What makes that happen ? Please help me, I want to know about them...

Thanks a lot for your answer
-Nina

*PS* Could Khp tracert the address above, copy and paste what response Khp gets ? Thanks ...

[khp]

Originally posted by hometown
Could you tell me why it doesnot follow the same route ? Why does that happen ? What makes that happen ? Please help me, I want to know about them...


Routing on the internet is not fixed, there is a degree of nondeterminism. Changes in routes might be cause at random (lets say there is 2 1Mbit routes between London and New York, it might be a good idear for the routers at either end to randomly switch between the two).
Route changes might also happen in response to changes on the network, lets say the most direct route beween two points becomes very slow or fails completly (either due to high trafic loads or a node failiure), the routers on the internet should then be able to route trafic around these trouble spots.

Originally posted by hometown
*PS* Could Khp tracert the address above, copy and paste what response Khp gets ? Thanks ...


traceroute to a1836.gb.akamai.net (194.255.208.110), 30 hops max, 38 byte packets
1 10.5.255.254 (10.5.255.254) 7.553 ms 67.922 ms 41.428 ms
2 10.255.250.57 (10.255.250.57) 8.307 ms 9.123 ms 9.322 ms
3 aarh-inet-gw.backbone.stofanet.dk (212.10.10.1) 10.707 ms 9.039 ms 9.344 ms
4 dk-hc-a2-geth2-0-access.hasselager.dk.telia.net (194.255.53.29) 9.480 ms 10.531 ms 9.500 ms
5 dk-ejb-b1-feth0-1-core.ejby.dk.telia.net (194.255.250.241) 75.657 ms 12.318 ms 13.410 ms
6 194.255.208.110 (194.255.208.110) 13.610 ms 13.565 ms 14.288 ms

[hometown]

Originally posted by khp
Routing on the internet is not fixed, there is a degree of nondeterminism. Changes in routes might be cause at random (lets say there is 2 1Mbit routes between London and New York, it might be a good idear for the routers at either end to randomly switch between the two).
Route changes might also happen in response to changes on the network, lets say the most direct route beween two points becomes very slow or fails completly (either due to high trafic loads or a node failiure), the routers on the internet should then be able to route trafic around these trouble spots.

traceroute to a1836.gb.akamai.net (194.255.208.110), 30 hops max, 38 byte packets
1 10.5.255.254 (10.5.255.254) 7.553 ms 67.922 ms 41.428 ms
2 10.255.250.57 (10.255.250.57) 8.307 ms 9.123 ms 9.322 ms
3 aarh-inet-gw.backbone.stofanet.dk (212.10.10.1) 10.707 ms 9.039 ms 9.344 ms
4 dk-hc-a2-geth2-0-access.hasselager.dk.telia.net (194.255.53.29) 9.480 ms 10.531 ms 9.500 ms
5 dk-ejb-b1-feth0-1-core.ejby.dk.telia.net (194.255.250.241) 75.657 ms 12.318 ms 13.410 ms
6 194.255.208.110 (194.255.208.110) 13.610 ms 13.565 ms 14.288 ms


Thanks a lot for all of those explanations :)
By the way, where did you learn those things ? I would like to know much like you....What books I should read ? Could you tell me ?

Once again, thanks a lot...

-Nina

[khp]

Originally posted by hometown
By the way, where did you learn those things ? I would like to know much like you....What books I should read ? Could you tell me ?


I've been studying computer sceince for just over 6 years, I was bound to pickup a few things here and there :).
I had a Distributed computing course, a couple of years ago, using these two books

http://www.cdk3.net
http://williamstallings.com/DCC6e.html

They are both very very dull if you ask me.

[hometown]

Originally posted by khp
I've been studying computer sceince for just over 6 years, I was bound to pickup a few things here and there :).
I had a Distributed computing course, a couple of years ago, using these two books
http://www.cdk3.net
http://williamstallings.com/DCC6e.html
They are both very very dull if you ask me.

:) Thank you so much...
-Nina

[hometown]

By the way, I have another question...
I got a firewall which I can see accessing IP from others, but how can I be sure if that IP is exactly from that site ?I meant how sure is sure enough ? How can i know that ? Or this is just an insane thought of mind because it is not possible and because there are so many an IPs come out after tracing ?
Thank a lot...
-Home

[khp]

Originally posted by hometown
I got a firewall which I can see accessing IP from others, but how can I be sure if that IP is exactly from that site ?


You can never be completly sure of anything. It is possible to make 'spoofed packages' with faked IPs using the right equipment. But doing that is of very little use, if someone send a request to your machine, he will usually expect to get a response, if he sends you a fake ip-address there is no way he can get the response, because your response would be sent to the fake ip instead of his ip.
So using spoofed packages with fake IPs is only usefull, for makeing denial of service attacks (trying to bring some site down, by flooding it with fake requests).

[hometown]

Originally posted by khp
You can never be completly sure of anything. It is possible to make 'spoofed packages' with faked IPs using the right equipment. But doing that is of very little use, if someone send a request to your machine, he will usually expect to get a response, if he sends you a fake ip-address there is no way he can get the response, because your response would be sent to the fake ip instead of his ip.
So using spoofed packages with fake IPs is only usefull, for makeing denial of service attacks (trying to bring some site down, by flooding it with fake requests).
I wont use them...
But actually I dont know how to use them though....:)

Thanks alot, you talk just like a ICP/IP specialist, I will compete with you 3 years later...

:)

[khp]

Originally posted by hometown
Thanks alot, you talk just like a ICP/IP specialist...

Me thinks you mean TCP/IP, but no, I'am no spcialist, I'am just a user with really big ears.

[hometown]

Originally posted by khp
Me thinks you mean TCP/IP, but no, I'am no spcialist, I'am just a user with really big ears.
:o You know, I just really didnot know about that typo...:)
Sorry, THANKs alot for your reminds:o

[I still canot be a specialist]:( Not yet...