[comp.protocols.tcp-ip] DNS Query server?

peters@jester.cc.msstate.edu (Frank W. Peters) (02/01/90)

Hello,

     Several months ago someone posted a offer of a very simple DNS resolver
server for UNIX.

     My nearest recollection of the operation of this software was that you
could connect to this code at a certain port (not the DNS port) and send a
name and receive an address.

     We now have a use for such a program.  Can anyone tell me where I might
get it?

Thanx
--Frank

Frank W. Peters        Systems Programmer     Computing Center & Services
peters@CC.MsState.Edu  Peters@MsState.Bitnet  (601)325-2942
"I can't give you brains, but I can give you a diploma." -- The Wizard of OZ

almquist@JESSICA.STANFORD.EDU ("Philip Almquist") (02/03/90)

Frank,
>      Several months ago someone posted a offer of a very simple DNS resolver
> server for UNIX.
>
>      My nearest recollection of the operation of this software was that you
> could connect to this code at a certain port (not the DNS port) and send a
> name and receive an address.

	I don't think that you are going to find what you're looking
for.  The UNIX DNS resolver is pretty simple.  The UNIX DNS server can't
be simple because it has a rather complicated task to perform.

	I suspect that what you heard about is something called an
IEN-116 name server.  IEN-116 was the predecessor to the DNS.  People
who have very obsolete Bridge terminal servers will have IEN-116
servers, or one was probably posted to comp.sources.unix at some point.

	Normally, the IEN-116 server merely turns around and asks a DNS
server, so unless there's some good reason why you have to use the
IEN-116 protocol instead of DNS directly this results in somewhat
pointless extra work.  If your system still has the old /etc/hosts
routines you can probably hack it pretty trivially to use the host table
instead (since that's the way the IEN-116 server originally worked), but
then of course you have to maintain the host table.

	Another problem with IEN-116 is that I've never seen a
publicly-available IEN-116 resolver, so you'd probably have to write
your own.  Since it's a different protocol than DNS, you need to change
rather more than just the port number in your DNS resolver to get it to
work.
							Philip

hsw@SPARTA.COM (Howard Weiss) (02/05/90)

Actually, I believe that what Frank Peters was looking for was
the service that host 128.218.1.109 provides on port 5555 - you
simply connect to that host at that port, type in a fully formed
internet host name and it responds with an internet address and
closes the connection.  I used it when I had a host that still
only had /etc/hosts and it did just what I needed - which was
basically a manual nslookup.

Howard Weiss

gnb@bby.oz.au (Gregory N. Bond) (02/07/90)

In article <9002022140.AA16279@jessica.Stanford.EDU> almquist@JESSICA.STANFORD.EDU ("Philip Almquist") writes:

	   Another problem with IEN-116 is that I've never seen a
   publicly-available IEN-116 resolver, so you'd probably have to write
   your own.  Since it's a different protocol than DNS, you need to change
   rather more than just the port number in your DNS resolver to get it to
   work.

Well, the Xylogics annex terminal server can also use an ien-116 name
service, and comes with source for an ien-116 daemon that is derived
from Berkeley software (it has a UCB copyright).  So check the
berkeley network sources.  (It is probably called named).  Also, most
4BSD-based systems come with an ien-116 server enabled in the inetd
configuration, so you may have it and just not know it.  (I know this
is true of SunOs 3.5 and 4.0).

Greg, who is in the process of installing an annex!
--
Gregory Bond, Burdett Buckeridge & Young Ltd, Melbourne, Australia
Internet: gnb@melba.bby.oz.au    non-MX: gnb%melba.bby.oz@uunet.uu.net
Uucp: {uunet,pyramid,ubc-cs,ukc,mcvax,prlb2,nttlab...}!munnari!melba.bby.oz!gnb

peters@CC.MsState.Edu (Frank W. Peters) (02/08/90)

In article <9002051330.AA00323@cheops.columbia.sparta.com> hsw@SPARTA.COM (Howard Weiss) writes:

   From: hsw@SPARTA.COM (Howard Weiss)
   Newsgroups: comp.protocols.tcp-ip
   Date: 5 Feb 90 13:30:23 GMT

   Actually, I believe that what Frank Peters was looking for was
   the service that host 128.218.1.109 provides on port 5555 - you
   simply connect to that host at that port, type in a fully formed
   internet host name and it responds with an internet address and
   closes the connection.  I used it when I had a host that still
   only had /etc/hosts and it did just what I needed - which was
   basically a manual nslookup.

Yes!  This is exactly what I want!  We're working on some networking
code on a UNISYS 1100 system that runs an operating system called
Exec8, doesn't have a C compiler or socket library and which has NO
usable hosts file system (as near as I understand it the hosts file is
compiled into the it's front-end-processors OS...so the FEP has to be
rebooted to add a new host to the table!!).

But enough about our troubles.  Where can I get a copy of the above
software?  the machine referenced above doesn't have an anonymous FTP
and the only machine at ucsf that I can find which does have anonFTP
(ccb.ucsf.edu) doesn't seem to have it either.

Any ideas?

--Frank

Frank W. Peters        Systems Programmer     Computing Center & Services
peters@CC.MsState.Edu  Peters@MsState.Bitnet  (601)325-2942
"I can't give you brains, but I can give you a diploma." -- The Wizard of OZ

postel@VENERA.ISI.EDU (02/09/90)

Hey!  

The IEN-116 name server is an old idea who's time is past.  
Kill and bury any code that still does IEN-116.

Please use the DNS instead!

--jon.

kre@cs.mu.oz.au (Robert Elz) (02/16/90)

From Jon Postel...
> The IEN-116 name server is an old idea who's time is past.  
> Kill and bury any code that still does IEN-116.

Isn't that a little harsh, IEN-116 is a trivial protocol
to implement, is very stable, and is ideal for proms, etc.

As long as the server does a DNS lookup (with addition of some
default domain if needed), what's the harm?

kre

wesommer@athena.mit.edu (Bill Sommerfeld) (02/16/90)

In article <3184@munnari.oz.au> kre@cs.mu.oz.au (Robert Elz) writes:

   Isn't that a little harsh, IEN-116 is a trivial protocol
   to implement, is very stable, and is ideal for proms, etc.

Not really; I've seen a simple DNS resolver which is under 1K of
object code (not counting a very primitive UDP/IP implementation).  It
*does* depend on knowing the address of another server which will
support queries with "recursion desired" set, but I suspect that the
IEN116 protocol has a similar requirement.

This isn't a full resolver -- it only knows how to translate hostnames
to addresses, but that's probably enough in most cases.

				- Bill

--
Henry Spencer is so much of a  |    Bill Sommerfeld at MIT/Project Athena
minimalist that I often forget |    sommerfeld@mit.edu
he's there - anonymous         |

dricejb@drilex.UUCP (Craig Jackson drilex1) (02/18/90)

The facts of IEN116 are: It's ridiculous for anything on the Internet.
It works fine for a small commercial Ethernet with a few PCs or 
terminal servers on it.

TCP/IP lives outside of the Internet.  Given the research/academic
orientation of the Internet, and the billing issues that exist,
commercial TCP/IP nets will likely continue to be isolated for quite
a long time.
-- 
Craig Jackson
dricejb@drilex.dri.mgh.com
{bbn,axiom,redsox,atexnet,ka3ovk}!drilex!{dricej,dricejb}

postel@VENERA.ISI.EDU (02/21/90)

 The facts of IEN116 are: It's ridiculous for anything on the Internet.
 It works fine for a small commercial Ethernet with a few PCs or 
 terminal servers on it.

 TCP/IP lives outside of the Internet.  Given the research/academic
 orientation of the Internet, and the billing issues that exist,
 commercial TCP/IP nets will likely continue to be isolated for quite
 a long time.


The notion that any network will remain small and isolated clearly ignores
all the experience of the last 20 years.

--jon.