Re: polygraph webaxe and HEAD

On Mon, 2002-08-12 at 14:36, Alex Rousskov wrote:
> On 12 Aug 2002, Robert Collins wrote:
>
> > I'm getting
> >
> > 554.16| CltXact.cc:442: error: 214/214 (c27) hit on reload request
> >
> > from the polygraph client during the webaxe-4 test.

> Polygraph uses
>
> Cache-control: no-cache
> Pragma: no-cache
>
> in the request, I think.

How does polygraph detect a hit? A magic cookie in the content?

If squid is checking that *all* the request headers are the same, and
joining the requests, then this seems ok to me from a RFC pov (the
content is still not stored). Does polygraph consider that OK?
 
> > (Yes, I know, too lazy to setup MIME logging just yet).
>
> Use "--dump errs" with polyclt to see more info. "--dump hdrs" will
> show headers for all transactions.
>
> HTH,

It does, thanks. I'll make some time soon to look into this.

Duane, do you recall this?
Henrik, is your rproxy branch the appropriate place to work on this?

Rob