Re: [squid-users] Setting up squid as an accel proxy

From: Martin Jacobson \(Jake\) <jake.jacobson@dont-contact.us>
Date: Fri, 21 Dec 2007 14:34:18 +0000 (GMT+00:00)

Amos,

Is the redirector requirement something new? We don't have it on our current Squid configuration which is running on version 2.2 STABLE 13.

Jake

Martin C. Jacobson (Jake)
Office of the DNI CIO
Intelligence Community Enterprise Services
Comm: (301) 317-7214 Secure: 369-6362

----- Original Message -----
From: "Amos Jeffries" <squid3@treenet.co.nz>
To: "Martin Jacobson (Jake)" <jake.jacobson@ugov.gov>
Cc: "Amos Jeffries" <squid3@treenet.co.nz>, "squid-users" <squid-users@squid-cache.org>
Sent: Thursday, December 20, 2007 6:26:01 PM (GMT-0500) America/New_York
Subject: Re: [squid-users] Setting up squid as an accel proxy

> Amos,
>
> I am taking a step or two back and try to explain what I want to do with
> Squid. I know this can be done because I am doing it with Squid version
> 2.2. I really want to upgrade to 2.6 but my configuration is just not
> working.
>
> I need to take my public URL which is a Linux server running squid. Squid
> is listening on port 80 for all in bound request and when it gets one it
> proxies the request to my back-end server which is a search engine. I
> don't want squid to do any caching of content, I just want it to proxy all
> web requests coming to www.jakesite.com.
>
> So in the above example, my public URL www.jakesite.com.
> My search engine's URL is search.jakesite.com:8000/
>
> When someone comes to www.jakesite.com with their browser, they should be
> presented the HTML search engine's page which is not on www.jakesite.com
> but is coming from search.jakesite.com:8000/
>
> I hope that this makes more sense. Again, thanks for your help.

It make sense and yes I understood that was what you want.
To do it you need a redirector which re-writes the URL

  http_port 80 accel defaultsite=www.jakesite.com

with a url_rewriter_program and location_rewriter_program

for their config settings see "OPTIONS FOR URL REWRITING" in
http://www.squid-cache.org/Versions/v2/2.6/cfgman/

Once that is done the URL are changed to squid retrieves the pages from
search.jakesite.com:8000 as any normal web request.

Amos

>
> Jake
>
> ----- Original Message -----
> From: "Amos Jeffries" <squid3@treenet.co.nz>
> To: "Martin Jacobson (Jake)" <jake.jacobson@ugov.gov>
> Cc: "squid-users" <squid-users@squid-cache.org>
> Sent: Thursday, December 20, 2007 3:40:00 AM (GMT-0500) America/New_York
> Subject: Re: [squid-users] Setting up squid as an accel proxy
>
> Martin Jacobson (Jake) wrote:
>> Amos,
>>
>> I have upgraded to STABLE17 but I still have the same issue.
>>
>> Basically I want to be able to go to URL in my browser like this:
>> http://linproxy1.mysite.com/. (linproxy1 is running squid)
>>
>> and have squid proxy the default search page from:
>> http://searchengine.mysite.com/
>
> Well, to CHANGE the URL like that you will need to use a redirector, not
> a peer.
> To peer as a forward-proxy squid simply imitates a server and has DNS
> pointed at it. The peer still needs to understand when requests for
> linproxy1.mysite.com arrive at it from squid through the private channel.
>
>>
>> Config settings:
>> http_port linproxy1.mysite.com:80 accel
>
> Could do with defaultsite=linproxy1.mysite.com to un-break some broken
> client software.
>
>> cache_peer 10.2.234.9 parent 80 0 no-query originserver
>> name=searchengine proxy-only
>
> proxy-only in accelerators does away with almost all of the actual
> benefit of 'acceleration' squid provides through caching unchanged
> content.
>
> Should work though as log as the peer knows its hosting
> inproxy1.mysite.com.
>
> Amos
>
>>
>>
>> Martin C. Jacobson (Jake)
>>
>> ----- Original Message -----
>> From: "Amos Jeffries" <squid3@treenet.co.nz>
>> To: "Martin Jacobson (Jake)" <jake.jacobson@ugov.gov>
>> Cc: "squid-users" <squid-users@squid-cache.org>
>> Sent: Tuesday, December 18, 2007 8:50:08 PM (GMT-0500) America/New_York
>> Subject: Re: [squid-users] Setting up squid as an accel proxy
>>
>>> I don't understand why I am having so much trouble getting something
>>> that
>>> seems to be so simple working. I have downloaded and installed
>>> squid-2.6.STABLE16 on my Linux box.
>>
>> Please use STABLE17, 16 has a major security bug.
>>
>>> I want to proxy my search engine's
>>> search page so I don't want squid caching the pages. I just want squid
>>> to
>>> act as a reverse proxy. After reading "Squid The Definitive Guide" and
>>> lots of posts on the web, I still can't get it to work. It would seem
>>> that the following basic config file should do the trick. This is
>>> based
>>> on http://wiki.squid-cache.org/SquidFaq/ReverseProxy
>>>
>>> http_port 80 accel defaultsite=linproxy1.mysite.com
>>> cache_peer searchengine.mysite.com parent 80 0 no-query originserver
>>> http_access allow all
>>>
>>> I can get squid to run with no errors, but it doesn't reverse proxy the
>>> search engine's search page. With my browser I can go directly to the
>>> search engine with no problems, but I get a "Connection has timed out"
>>> error when I try to go through linproxy1.
>>>
>>
>> So http://linproxy1.mysite.com is the publicly accessible website for
>> the
>> search engine?
>>
>> Does the machine at searchengine.mysite.com serve pages for
>> http://linproxy1.mysite.com properly?
>>
>> Amos
>>
>
>
> --
> Please use Squid 2.6STABLE17 or 3.0STABLE1.
> There are serious security advisories out on all earlier releases.
>
Received on Fri Dec 21 2007 - 07:34:40 MST

This archive was generated by hypermail pre-2.1.9 : Tue Jan 01 2008 - 12:00:02 MST