Tracking

In search of the elusive accurate stat.

[Murrayson]

Hi all,
Not sure where to post this, here seems most appropriate.

My question is this:

Has anyone tried the “Meter” header (rfc2227), and is it an accepted HTTP1.1 extension?

We have been having huge problems with all the transparent proxies out there and would like to build a tracking mechanism to use extended (server side) logging and compare it to custom (client-side) image logging.

Theoretically client-side logs should equal server-side logs + meter reports. We have an unusually high no of internal refs. Points to those damn caches!! It would be nice to believe that our log stats are at least faintly plausible.

[Brett_Tabke]

Forgive me for being a lazy searcher, but have you got a handy url for rfc2227? This sounds new to me. I normally use a combo of server side generated logs and cgi image calls that are not cached by browsers. That gives you nearly 100% coverage.

[Murrayson]

Hi Brett,

[eugene.kashpureff.org...]

You can also find all the rfc's @ www.w3.org all the standards. Not that everyone adheres to them. But the big boys do to a large degree, I hope. Just have to find time to read them ..LOL

I see that the HTTP1.1 rfc 2068 has been made redundent my rfc 2616.

The 2227 rfc deals with extentions to Http1.1 ... trying to deal with the caching problem.

[cisco.com...] is also a good article .. great ref material @ the bottom.

Later.Brett_Tabke

[bartek]

Hi Brett,

"...cgi image calls that are not cached by browsers."

Can you expand on this a bit?