Re: [SQU] Host header based Acceleration?

The added delay isn't noticeable. Squid starts several instances
of your redirector script when it initializes. Each of the redirectors
sits in an endless loop, getting URLS and spitting back the rewritten
URLs (or the same string if it doesn't match one of the search patterns)
until Squid exits.

If on the other hand Squid had to start up a new redirector instance for
each url, the delay would be much larger, but the way it is actually
handled is very efficient.

So, it takes a bit of RAM, but the added delay is very low.

Mike

> Won't this be inefficient as for each URL squid will have to rewrite
> it to another? Is there no another way within squid's acceleration
> features, without introducting a separate redirector program, that
> based on host header A squid will automatically pull the web site from
> server1 :80 and based on host header B, pull the site from
> server1:81?

> Thanks for any help

> On Wed, 25 Oct 2000 11:11:24 -0500 (CDT), you wrote:

> >> Can you set up squid to pull web pages based on host headers.

> >> If requests www.x.com ----> pulls from server1 on port 80
> >> www.y.com -----> pulls from server 1 on port 81

> >You should be able to do this with a redirector script -
> >rewrite www.x.com to server1:80 and www.y.com to server1:81

> >Mike

--
To unsubscribe, see http://www.squid-cache.org/mailing-lists.html