Is going to "http://somewhere.com." faster?
Mark_Andrews at isc.org
Mark_Andrews at isc.org
Thu Jul 29 23:29:08 UTC 1999
Actually, the behaviour is usually the same. Nslookup's
behaviour has changed over the years to match that of the
resolver.
The resolver (and nslookup) behaviour was changed years
ago to address the problems identified in RFC1535. Ask
you vendor for patches to address the issues identified
in RFC1535.
Mark
RFC1535:
A Security Problem and Proposed Correction With
Widely Deployed DNS Software, E. Gavron, ACES Research
Inc., October 1993.
> You don't need a dot at the end in the browser. NSLOOKUP behavior is not
> usually the same a resolver. Resolvers will generally NOT append the
> local domain to a name that contains any dots in it. So, your browser
> will not append mycompany.com to request that have a dot in the name
> already. It doesn't need the ending dot. nslookup is a useful tool, but
> do not assume it behaves exactly like the local resolver.
>
> Michael Voight
> CSE, Cisco TAC
>
> yong321 at yahoo.com wrote:
> >
> > I'm learning DNS. At the nslookup> prompt, I "set debug". Then query
> > "www.uh.edu" for University of Houston (I work at Houston). It first
> > tries to find "www.uh.edu.mycompany.com" and gets 0 answers. Then it
> > tries to find "www.uh.edu" and finds it. Now if I type "www.uh.edu."
> > (with the appending dot), it finds it with one attempt. This is easy to
> > understand.
> >
> > Then I think if I go to the homepage of any Web site (not inside the
> > site), I can append a dot to the host name to speed up DNS lookup a
> > little bit. So in the browser, instead of typing the URL "www.uh.edu"
> > (or "http://www.uh.edu") I type "www.uh.edu.". Please comment. Thanks.
> >
> > Yong
> > Email:yong321 at yahoo.com
> >
> > Sent via Deja.com http://www.deja.com/
> > Share what you know. Learn what you don't.
>
>
--
Mark Andrews, Internet Software Consortium
1 Seymour St., Dundas Valley, NSW 2117, Australia
PHONE: +61 2 9871 4742 INTERNET: marka at isc.org
More information about the bind-users
mailing list