A and H root-servers.net giving out bad information?
Joseph S D Yao
jsdy at cospo.osis.gov
Mon Sep 11 22:46:16 UTC 2000
On Mon, Sep 11, 2000 at 03:22:48AM -0700, Joe Wagner wrote:
>
> Hello,
> Thanks in advance for your help -- and your patience. For a while
> now I have been having periodic problems getting through to a mail
> server at Stanford University, though no one else complained of any
> difficulties. A couple weeks ago I finally noticed that when those
> problems occurred, my email program(s) were using an old and now bad IP
> address (36.93.0.31) rather than the new one (171.64.48.31) for
> cdr.stanford.edu.
> More than a few moons ago IIRC stanford.edu made the final switch
> from using all of 36.x.x.x to smaller hunk of the IP space in
> 171.x.x.x. I could not figure out where and why my name servers were
> getting the old IP address -- until tonight: For some reason
> A.root-servers.net (and H) is giving out the now bogus IP address. Thus
> every so often when my DNS server happen to decide to go to A or H
> rather than another root-server, it caches a bogus value until it
> expires, or I notice and flush the cache.
> Is there something special about A and H that make them cling to
> old records? For that matter, if this isn't too a dumb question, why
> would A and H give an answer rather than just an Authoritative Record
> like the rest of the root-servers?
>
> Thank you again,
>
> Joe
>
> PS, below are the results of a couple digs. Note that I think in the
> past cdr.stanford.edu ran a name server though it doesn't anymore.
The last is the clew. There is probably a residual Host record with
the old address. You will have to get the stanford.edu hostmasters to
either fix this or (better) delete it.
--
Joe Yao jsdy at cospo.osis.gov - Joseph S. D. Yao
COSPO/OSIS Computer Support EMT-B
-----------------------------------------------------------------------
This message is not an official statement of COSPO policies.
More information about the bind-users
mailing list