Re: [squid-users] acl forbidden_domains dstdom_regex "file.txt" with huge file fails

From: Marcus Kool <marcus.kool_at_urlfilterdb.com>
Date: Wed, 20 Jun 2012 11:54:29 -0300

On 06/20/2012 06:43 AM, Matus UHLAR - fantomas wrote:
> On 19.06.12 18:52, Stefan Bauer wrote:
>> with a 30 MB file. Squid is instantly terminating if this acl-stanza is set active. Where can and do we have to tune squid settings to achive this?
>
> terminating with what reason? I would not wonder if all the regexed would not fit into the memory. Note that using bunch of regexes slows squid down very much - it has to compare tons of regexps for
> every request.
>
>> We're aware of third-party software like squidguard for this task - we only want to use additional software if everything else fails :)
>
> well, now squid failed...

ufdbGuard is much faster than squidguard with regular expressions since it optimises a set of regular expressions into one regular expression.
The regex man page says there is no limit to a regular expression other than limits imposes by hardware or the kernel.

Marcus
Received on Wed Jun 20 2012 - 14:54:34 MDT

This archive was generated by hypermail 2.2.0 : Thu Jun 21 2012 - 12:00:03 MDT