help - no DNS entry

Michael Huber mhuber1 at umbc.edu
Tue Jun 10 10:27:30 EDT 1997


On Tue, 10 Jun 1997, Peter Murray wrote:

> "DNS" is an acronym for "Domain Name Service".  TCP/IP (and consequently the
> Internet) is made up of connection between a bunch of numbers (called an IP
> address; like "129.22.138.1") and DNS provides (among other things) a mapping
> of human-rememberable names to these numbers.  For instance, people send mail
> to you at a machine called email.mesd.k12.or.us, but machine actually has the
> address "198.236.66.3", and any machine sending your machine mail must
> translate the name to the address in order to find your machine on the
> network.  The same sort of thing happens in the web:  "www.cwru.edu" is
> translated to an IP address by the DNS, which the browser then uses to
> connnect to our web server.

Try the following, assuming you have access to the Unix 'shell/command
prompt account' with the following programs installed:

1) nslookup abc.def.ghi.com

   this should return the IP Address of the site in question and you can
   try using that in place of the name in the URL.  If DNS lookups are
   failing regularly, this will bypass the lookup by using the IP address
   directly.

2) traceroute abc.def.ghi.com

   this will give you a list of the site the request passes through to get
   to your destination.   this may tell you where your request is failing
   and whether it your problem, their problem or someone in betweens.

3) ping abc.def.ghi.com

   this will tell if packets (requests) are getting through, how long it
   takes and what percentage are getting lost.

In short if I (from lib.umbc.edu) enter a URL of www.yahoo.com (X) the
following will happen:

1) baggins (my computer) doesn't know where X is so it

2) sends the request to the DNS at umbc.edu which

3) sends it to the DNS at .edu (root of edu) which

4) sends it to the DNS at the root of the world which

5) sends it to the DNS at .com (root of com) which

6) and so on...

A breakdown in any of these links or simply taking too long to respond can
cause the DNS query to fail.

Most of the time if the query fails immediatly, the URL is mistyped in the
hostname portion or there is a problem with one of the machines in the
link.  Hope this helps.



More information about the Web4lib mailing list