Re: Taxonomy Term closure needed

From: Joe Touch (touch@ISI.EDU)
Date: Mon Sep 13 1999 - 14:37:55 MDT


> Date: Mon, 13 Sep 1999 20:17:17 +0200
> Subject: Re: Taxonomy Term closure needed
> To: Joe Touch <touch@ISI.EDU>
> From: Henrik Nordstrom <hno@hem.passagen.se>
> Cc: wrec@cs.utk.edu
>
> Joe Touch wrote:
>
> > What concerns me about the above is the use of "intent"
> > in a definition. There must be some objective metric.
> > Management control, service agreeements, etc. aren't
> > useful objective metrics.
>
...
> > Is there any definition of a reverse proxy that is objective?
> > (and can distinguish it completely from a non-reverse proxy cache?)
>
> What about this:
>
> A reverse proxy is officially published (in DNS and/or by official IP
> address) as the origin server and forwards received request towards the
> real origin server. Common usage is relaying requests through a server
> side firewall or providing alternate access points where traffic is
> relayed through private network(s).

Why is it useful or necessary to distinguish what port a proxy sits on?

I saw the discussion that distinguishes whether a proxy is
explicitly forwarded to or in-place-of the server itself. While
this is a clean definition, why make the distinction? Are there
different requests/responses?

(sure, there are different ways to tune things, but that's a function
of the _data_, not specifically where the proxy is).

Do all requests out of a "server-side proxy" always fall-through
to a single address, rather than 'anywhere in the network'?

Here are the distinctions that may be possible:

        - conventional proxies and clients do not explicitly
          direct requests to a server-side proxy cache
          (the GET request address matches the TCP connection address)

                is this true?

        - requests missing in a server-side cache fall through
          only to the origin server (i.e., a single fixed IP address
          and port); they are not forwarded elsewhere

                  is this true?

Even if both are true, these may not be meaningful differences.
The only meaningful difference may be whether the protocol
changes - does it?

Joe



This archive was generated by hypermail 2b29 : Thu Nov 18 2004 - 11:21:27 MST