Re: Hi there

From: Alex Rousskov <rousskov@dont-contact.us>
Date: Mon, 15 Apr 2002 08:42:46 -0600 (MDT)

On Sun, 14 Apr 2002, Yee Man Chan wrote:

>
> I just asked one ICAP person. He says ICAP doesn't
> address the questions I had. He suggested me to go to
> OPES for more details. http://www.ietf-opes.org

I did not say ICAP addresses these questions. I just answered your
specific questions based on my understanding of how an ICAP server can
be used for blocking or modifying the content. Sure, there are plenty
of other possible modes of operation here, including ICAP server being
in some relationship with the origin server.

Alex.

> >
> > Squid is configured with an ICAP server IP:port,
> > similar to the
> > cache_peer option. Content_filter?
> >
> > Multiple ICAP servers and ACL-based redirection need
> > to be supported
> > as well.
> >
>
> I agree this should be the approach when we are
> dealing with a content adaptation server (CAP) that
> doesn't have a close tie to the original web server.
> For example, the CAPs that does translation, virus
> scan, etc. This is also the approach adopted by
> Geetha's patch.
>
> But for the CAPs that modifies the content or the ones
> that map different client to different URL, e.g.
> i-mode URL for cell phone client, another URL for PDA,
> they have closer tie to the original server. Therefore
> I would think some kind of interaction between
> original server, CAP and squid will have to be in
> place.
>
> >
> > This can be a result of a matching ACL rule in Squid
> > and/or a result
> > of an internal ICAP server configuration.
> >
>
> Again, this depends on the closeness of the CAP to the
> original server.
>
> Yee Man
>
> > Alex.
Received on Mon Apr 15 2002 - 08:42:48 MDT

This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:15:10 MST