Re: [squid-users] Re: Squid3 Reverse Proxy based on url path

From: Amos Jeffries <squid3_at_treenet.co.nz>
Date: Fri, 14 May 2010 10:21:23 +1200

rainolf wrote:
> i saw the config example in squid-cache.org :
>
> For our example here the websites /foo directory alone is hosted on a second
> server:
>
> * example.com is hosted on server1
> * example.com/foo is hosted on server2
>
> acl foo urlpath_regex ^/foo
>
> cache_peer ip.of.server1 parent 80 0 no-query originserver name=server1
> cache_peer_access server1 deny foo
>
> cache_peer ip.of.server2 parent 80 0 no-query originserver name=server2
> cache_peer_access server2 allow foo
> cache_peer_access server2 deny all
>
> this is not my case cause foo directory is not present on my 2nd
> webserver..its only a fake directory just to route requests on 1st or 2nd
> server depending on fake url path..
>
> xxx.domain.com/a ------------> redirect to server1.domain.local
> xxx.domain.com/b ------------> redirect to server2.domain.local
>

First you asked how to switch based on path prefix. Which is the answers
you got.

Now you talk about rotating between two source peers with the same content.

These are different topologies with very different config. As you have
noticed the directory split configuration will break a same-content
configuration.

What are you actually trying to accomplish?

Right now it sounds like you have two identical peers and want to load
balance between them?

Amos

-- 
Please be using
   Current Stable Squid 2.7.STABLE9 or 3.1.3
Received on Thu May 13 2010 - 22:21:33 MDT

This archive was generated by hypermail 2.2.0 : Fri May 14 2010 - 12:00:06 MDT