[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]


To: Daniel Senie <dts@senie.com>
Cc: Robert Elz <kre@munnari.OZ.AU>, ngtrans@sunroof.eng.sun.com, namedroppers@ops.ietf.org, ipng@sunroof.eng.sun.com, dnsop@cafax.se
From: David Terrell <dbt@meat.net>
Date: Wed, 8 Aug 2001 18:29:54 -0700
Content-Disposition: inline
In-Reply-To: <5.1.0.14.2.20010808120719.03df9bd0@mail.amaranth.net>; from dts@senie.com on Wed, Aug 08, 2001 at 12:12:15PM -0400
Reply-To: David Terrell <dbt@meat.net>
Sender: owner-dnsop@cafax.se
User-Agent: Mutt/1.2.4i
Subject: Re: (ngtrans) Joint DNSEXT & NGTRANS summary

On Wed, Aug 08, 2001 at 12:12:15PM -0400, Daniel Senie wrote:
> At 11:44 AM 8/8/01, Robert Elz wrote:
> >     Date:        Wed, 08 Aug 2001 11:13:52 -0400
> >     From:        Daniel Senie <dts@senie.com>
> >     Message-ID:  <5.1.0.14.2.20010808111308.03f697d0@mail.amaranth.net>
> >
> >   | Not to mention melting the 'net under ever increasing DNS load, since 
> > we'd
> >   | no longer be able to cache anything.
> >
> >Huh???
> >
> >No-one ever said anything about changing the definition or use of the
> >TTL field in DNS replies.   If you get a TTL that says an address is
> >valid for a day, then you can keep using it for a day without checking
> >again.  Or you can check again every 5 minutes if you want to, but
> >the answers will just keep coming back from your local cache, each with
> >a TTL 5 minutes shorter than the previous time...
> 
> Reread what Keith wrote. If applications are going to use DNS to check for 
> changes in addressing, how is caching going to help? You're suggesting the 
> local caches just answer the every-5-minute lookups, but that's useless if 
> the DNS lookups are used as a part of multihoming. I interpreted the 
> periodic lookup as being a way for applications to find out that a remote 
> machine has migrated to a new address. If local caches mask that migration, 
> how's that help?
> 
> Multihoming has to involve resiliancy. If the addresses are cached for a 
> day, saying "oh, your application will start working again tomorrow" is 
> unlikely to cut it.

I think they're talking about reestablishing existing connections
if the address published in the DNS changes.

I think that's pretty silly.  Application protocols should (where
appropriate) be able to reconnect, or users can -- and DNS records
near a renumbering event should have low TTLs, or multiple A* records
for a multihomed situation, and applications should not be caching
records excessively (or at all, really), and making multiple attempts
at multiple A* records.

-- 
David Terrell           | "We must go forward, not backwards; upwards,
Nebcorp Prime Minister  | not forwards; and always twirling, twirling,
dbt@meat.net            | twirling towards freedom!"
http://wwn.nebcorp.com/ |  - The Simpsons

Home | Date list | Subject list