Re: [squid-users] i'm getting a lot of output in my cache.log about status_code acl

From: Eliezer Croitoru <eliezer_at_ngtech.co.il>
Date: Tue, 12 Jun 2012 07:47:19 +0300

<SNIP>>
>
> Cool. Are you able to write this up into the YouTube wiki page?
>
>
> Amos
>
i will need to get a test page first to feel the "wiki" because i have
never written one.
also my user from a reason wasnt activated.
i didnt got the mail (my server is sitting on a dynip server so it might
be it but it seems ood).

the thing is that it can be done also with a url rewriter almost the
same with icap.
the only problem is that you cant do the trick of icap on response.
but as i'm thinking on it it can be done with "reply_header_replace"

the idea that can be implemented wiht url_rewrite and two squid
instances is:
first url rewriter psudo:
strips that dynamic data into memory
compose an internal url with the data.
writes the internal url and original url in a memory database.
send squid the new url..
squid has acls for the internal domain to get it using the cache peer...

second url rewriter psudo:
strips the dynamic content data\use the url as id
check in the memory db if exists and if so grab the original url and
send it to squid.

client - asks yt
    |
    |
squid1 asks from squid 2 yt.squid.internal (client dont know about it)
    |
    | ( mysql memory db for coordination of acrobatics ;) )
    |
squid2 gets from yt the original url (squid1 dosnt know about it )
    | here on squid2 i can manipulate the headers rewriting
   \|/

    YT

well it's not the store_url_rewrite but acrobatics so i'm kind of proud
of it.

will try the header rewriting on squid to make the whole thing more
perfect and to even work without icap.

Eliezer

-- 
Eliezer Croitoru
https://www1.ngtech.co.il
IT consulting for Nonprofit organizations
eliezer <at> ngtech.co.il
Received on Tue Jun 12 2012 - 04:47:26 MDT

This archive was generated by hypermail 2.2.0 : Tue Jun 12 2012 - 12:00:03 MDT