Looking for DNS tracing tool

William Stacey [MVP] staceyw at mvps.org
Mon Jul 15 13:11:10 UTC 2002


Might try to "wrap" dig in your own script or program and do the last lookup
via code if record returned is a CNAME.  If its not a CNAME, then (at least
on my system) dig returns the A record or MX, etc.

--
William Stacey, MCSE
Windows Server MVP



<phil-news-nospam at ipal.net> wrote in message
news:agmv71$bg4g$1 at isrv4.isc.org...
> I'm looking for a DNS tracing tool that doesn't stop and NS records,
> but follows through and completes the trace, not only following
> delegations all the way, but also getting the final record and show
> it, PLUS, if it is a CNAME and something other than CNAME was requested,
> it should follow the CNAME as well, exactly like any recursive lookup
> would do, and shows everything down to the final record, including it.
> So if I'm doing a A record lookup for a name which has a couple of CNAME
> references, it will follow through both CNAME references and give the
> final A record (and everything in between).  Basically it would be a
> full recursive trace.
>
> I've tried "dig +trace" and a program called dnstracer, but both of
> the stop at the last name server and won't finish, and both fail to
> follow a CNAME.
>
> An added plus, which dnstracer can do somewhat, is to do a "full spread"
> recursion, where ever reference is followed through at every level, even
> if they go different ways.  For example, every name server is queried
> for multiple NS records.  And if those servers have different data that
> would recurse in different directions, all of them would be followed.
>
> It would also be nice if it encounters obvious errors, to note the errors
> and keep on working.
>
> Another added plus would be to follow PTR records to corresponding A
recors.
>
> --
> -----------------------------------------------------------------
> | Phil Howard - KA9WGN |   Dallas   | http://linuxhomepage.com/ |
> | phil-nospam at ipal.net | Texas, USA | http://phil.ipal.org/     |
> -----------------------------------------------------------------
>




More information about the bind-users mailing list