[squid-users] Re: Problem with Squid 2.6 as reverse proxy

From: Santiago Del Castillo <santiago.del.castillo@dont-contact.us>
Date: Thu, 7 Jun 2007 20:12:30 -0300

That didn't work well

I used \([1-9]|[1-4][0-9]|50\) and saw:

2007/06/07 18:23:10| Failed to select source for
'http://www2.xxxxxx.com/2/AF/AA/Sol/last_photo.jpg'
2007/06/07 18:23:10| always_direct = 0
2007/06/07 18:23:10| never_direct = 0
2007/06/07 18:23:10| timedout = 0

The pattern I found is that it just fails with units ([1-9]) the rest
works OK. What it could be?

Tried to use [123456789] and it also fails.

Cheers,
Santiago

On 6/7/07, Henrik Nordstrom <henrik@henriknordstrom.net> wrote:
> ons 2007-06-06 klockan 19:34 -0300 skrev Santiago Del Castillo:
> > Hi henrik,
> >
> > One question: can wildcards be used on cache_peer_access?? Because
> > i've 100 domains (www1.example.com, www2.example.com ...
> > www*.example.com) forwarded to one specific origin server and it could
> > be great if i could use www*.example.com on cache_peer_access rule
>
> Yes, using the dstdom_regex acl in cache_peer_access. Or if it's the
> whole domain then use a dstdomain acl .example.com
>
> > Also that may change and i've to forward from www1 to www50 to other
> > origin server... How should I do that in a few lines and not more than
> > 50?
>
> 1 to 50 is ([1-9]|[1-4][0-9]|50) in regex, so www(1-50).example.com
> becomes
>
> ^www\([1-9]|[1-4][0-9]|50\)\.example\.com$
>
> But you can also use the dstdomain acl, with a list of all domains. Or
> exclusions using !acl in cache_peer_access.
>
> Regards
> Henrik
>
>
Received on Thu Jun 07 2007 - 17:12:34 MDT

This archive was generated by hypermail pre-2.1.9 : Sun Jul 01 2007 - 12:00:04 MDT