Subject: Re: sethostent(1): is it really that useful with a DNS?
To: Perry E. Metzger <perry@piermont.com>
From: Chuck Cranor <chuck@research.att.com>
List: tech-userlevel
Date: 06/17/2001 12:46:12
On Sun, Jun 17, 2001 at 12:22:34PM -0400, Perry E. Metzger wrote:
> Chuck Cranor <chuck@research.att.com> writes:
> > in the typical config (local DNS "close" to resolver client on network)
> > i don't think you'll get any real performance boost from TCP.   e.g. 
> > consider NFS which uses UDP in most cases like this.
 
> Actually, you're almost always better off with TCP these days on
> virtually any protocol. The "overhead" of TCP turns out to be more
> than offset by getting real congestion control. An extremely unloaded
> LAN or something where you have very small transactions are the only
> (possible) exceptions.

yes, that's why i said: 'in the typical config (local DNS "close" to 
			resolver client on network)'

i contend that DNS transactions are relatively small and most 
LANs between resolver clients and their local DNS servers are 
uncongested.


chuck