Dear Noel,
> I have been visiting the official Squid site and several other related
> sites but I haven't found any real documentation such as one would find
> for INN or sendmail.
> I'm looking for something on the lines of design model, communications model,
> cacheing algorithms, etc. Am I perhaps not fully aware of all the available
> info. ?
>
Have a look at URL: http://www.uninett.no/prosjekt/desire/arneberg/
It's a Web caching architecture document we have written in the framework
of the Desire/Web-caching Project..
> Also, ASAIK, if you got a cache hierarchy, such as a parent and a single
> sibling, where all the clients are going through the sibling, on a miss,
> the sibling will interrogate the parent. If the parent can't deliver the
> object, the sibling establishes an Internet connection itself (pls. see
> diagram).
>
>
> client --------------------------> sibling
> ------------------------------------> Internet
> |
> |
> \ /
> parent
>
>
Your so called "sibling host" (say host A) is in fact the first-level Web
caching server for the browsers. This Web caching server can have a sibling
relation with your so called "parent host" (say host B). Host A will send
an ICP request to host B to check if the requested object is available.
If so, the parent send it imediately back (if small enough) in the ACK-ICP
packet or send a empty ACK packet so that host B can fetch it with a GET
request. If the requested object is not on host B, it will send a NACK to
Host A which will fetch the object from the originating Web server..
It's also possible
that host A is a child of host B. In this case the requested object will
be fetched from the originating Web server (or from other sibling/parent
Web caching servers) by Host B.
> Would it be possible for the parent to retrieve the object itself, and then
> forwards it to the sibling, both servers having the opporunity to cache the
> object ?
>
Only if Host B is the parent of host A.. A sibling host will not ever fetsh
an object for another sibling relation from the internet..
Cheers, Henny
-- ----------------------------------------------------------------------------- E-Mail: Henny.Bekker@cc.ruu.nl ! Disclaimer: The main obstacle to progress http http://www.ruu.nl/~henny ! progress is not ignorance,but PTT: Voice: +31 30 2536971 Fax: +31 30 2531633 ! the illusion of knowledge X400: /G=Henny/S=Bekker/OU=cc/O=ruu/PRMD=surf/ADMD=400net/C=nl o Paper: H.J.Bekker, Utrecht University, Computer Centre _ /- _ Po Box 80011, 3508 TA Utrecht Nederland (_) > (_) -----------------------------------------------------------------------------Received on Thu May 29 1997 - 16:21:07 MDT
This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:35:17 MST