Re: [squid-users] Selective Logging

>Shouldn't be too difficult to make a simple perl script to pull out only
>external requests. As far as I know, without attacking the Squid code,
>you won't be able to use ACLs for logging decisions. A perl script
>would be a lot easier, and work just fine.

The only problem I have with this is ownership of the file. I run Squid with
an effective user of squid, not root. When I run the file through a
filtering script either as root, or even su'd to squid, the ownership of the
access log changes to root, and squid throws it's teddies out of it's cot
because it then can't write to the file. I appreciate that I can then change
the ownership of the file back to squid upon completion, but this is not
only a pain, but runs the risk of missing log entries while the filter runs.
Also, as log files grow, the size of the file this script has to deal with
grows to a large size. I run webalizer on the stats, and gather a month's
log at a time. By the end of the month, the log file can be huge, and any
processing of it is likely to take a while, meaning that more hits are
likely to be lost. I would prefer not to upset Squid more than is neccessary
too...

Is this the only way, or does anybody have any further comments/suggestions?

As ever, any thoughts are appreciated.

Simon

------------------------------------------------------------
Simon Allenby - IT - Project Telecom
email - simona@projtel.co.uk
Direct line 01636 615200

 ------------------------------------------------------------
 This communication is confidential and is intended
 only for the person to whom it is addressed.
 If you are not that person you are not permitted to
 make use of the information and you are requested
 to destroy the copy in your possession immediately.
 ------------------------------------------------------------
Received on Tue Jun 05 2001 - 03:04:37 MDT