W3C home > Mailing lists > Public > ietf-http-wg-old@w3.org > January to April 1998

Re: Hit-metering implementations

From: Jeffrey Mogul <mogul@pa.dec.com>
Date: Mon, 05 Jan 98 11:06:20 PST
Message-Id: <9801051906.AA20918@acetes.pa.dec.com>
To: Jayakumar Ramalingam <JAYAKUMAR@novell.com>
Cc: http-wg%cuckoo.hpl.hp.com@hplb.hpl.hp.com, JAN_PROVAN@novell.com, KNEWSOM@novell.com, MKUMAR@novell.com
Jayakumar Ramalingam writes:

    Please redirect me, if this isn't the right forum for this
    question.

Since RFC2227 is a product of this working group, this is the right
forum for the question.  (Sorry for the slow response; the mailing
list was shut down for a few weeks, and I wanted to wait until it
revived before responding.)

    I am looking for Web server implementations of Hit-metering for
    testing against our Proxy server implementation.

    I tried searching the working group archive for such information
    and I could not find any announcements regarding implementation of
    RFC 2227.

RFC2227 is a "Proposed Standard" of the IETF.  Before it can progress
to "Draft Standard" status, we will need to demonstrate the existence
of at least two independent and interoperable implementations.  It's
not entirely clear if that means one origin-server implementation plus
one proxy-cache implementation, or if we need two of each.

At any rate, I am aware of several people working on proxy implementations.
It would (as Jayakumar has noted) be nice to have some server
implementations to test these against.  If anyone is working on
an origin-server implementation of RFC 2227, even in an experimental
form, please either post a message to this mailing list, or (if you
are concerned about too much publicity) send email to me, and I will
serve as a match-maker for implementation testers.

Note that the tricky part of an origin-server implementation is probably
not the RFC2227 protocol per se, but finding the right way to record
the use-counts in a useful way (e.g., extending the server log file
format).  However, this problem is NOT part of the protocol specification,
and I do not believe that an implementor working on solving this problem
is required to reveal the solution as part of the IETF-mandated
interoperability testing.  I.e., if you are worried about revealing
your proprietary solution to the records-keeping problem, we can work
something out to allow protocol interoperability testing while preserving
your proprietary information.

-Jeff
Received on Monday, 5 January 1998 11:07:51 EST

This archive was generated by hypermail pre-2.1.9 : Wednesday, 24 September 2003 06:33:09 EDT