W3C home > Mailing lists > Public > ietf-http-wg-old@w3.org > September to December 1996

Hit Metering - report of 0/0

From: Patrick McManus <mcmanus@appliedtheory.com>
Date: Thu, 21 Nov 1996 13:49:07 -0500 (EST)
Message-Id: <199611211849.NAA06700@pat.appliedtheory.com>
To: http working group <http-wg%cuckoo.hpl.hp.com@hplb.hpl.hp.com>
Folks,

I've got a comment on draft-mogul-http-hit-metering-00.txt

>   A proxy SHOULD NOT transmit "Meter: count=0/0", since this conveys no
>   useful information.

I believe that this draft should be silent on this issue, largely
because whether a count=0/0 response is useful information or not is a
context sensitive question.

The origin server is explicitly given the right the right to decide
whether or not to enter into a caching 'contract' with a proxy based
upon any criteria it chooses. I can envision a situation where
reporting of 0 hit counts would be a valuable criteria to base this
decision on.

Consider a server that is very concerned about receiving accurate hit
counts. This proposal requires proxies that agree to return hit counts
to return them with a 'best-effort' (which I think is the appropriate
requirement) but doesn't guarantee their return in the case of network
or system failure. If the server in question has reason to believe
that the proxy's 'best-effort' isn't sufficient it may choose to
answer every request with a "cache-control: proxy-revalidate". I think
that one criteria for determining a proxy's success with 'best-effort'
would be that proxy's history of returning hit counts in the past that
it was committed to returning to this server. (for example this
heuristic could be used to judge network reliability between the two
machines).. a report of 0/0 when an unused entry is flushed from the
proxy's cache still serves as a sense of closure and completion of
contract to the origin server and _could_ be useful in future decision
making. Mind you I don't think that reporting 0/0 cases should be
required, or even necessarily encouraged, but discouraging them limits
specific options that I think should be implementation issues.

-Patrick

--
Patrick R. McManus - Applied Theory Communications -	Software Engineering
http://pat.appliedtheory.com/~mcmanus			Programmer Analyst
mcmanus@AppliedTheory.com	'Prince of Pollywood'	Standards, today!
*** - You Kill Nostalgia, Xenophobic Fears. It's Now or Neverland. - ***
Received on Thursday, 21 November 1996 17:02:17 EST

This archive was generated by hypermail pre-2.1.9 : Wednesday, 24 September 2003 06:32:16 EDT