Re: Correct to log query args with URL?

[Randy Terbush <ra...@zyzzyva.com>]

> 
> > 
> > I just noticed that we are logging the query arguments when loggin a URL.
> > Is this correct?
>  
> 
> It's standard CLF practise to do so isn't it ?, or are you
> talking about something other than QUERY_STRING ?

No, I *am* talking about QUERY_STRING.  This seems kind of ..er
*stupid*....

Anyway, in case anyones interested, I hacked a new version of the
logger, that is *not* configurable, but suits my needs a bit better.
I am under the gun to start getting some accounting in place and
am taking the short route. I wrote a script to convert from CLF to
my new format and just realized that the QUERY_STRING is being logged
as part of the request in CLF.  *sigh*  

I have ditched the idea of talking directly to a database or logging
daemon.  It creates way to fragile an environment to depend on.
Using the filesystem to cache this is much safer.

Re: Correct to log query args with URL?

[Brian Behlendorf <br...@organic.com>]

On Sun, 23 Jul 1995, Randy Terbush wrote:
> > > I just noticed that we are logging the query arguments when loggin a URL.
> > > Is this correct?
> >  
> > 
> > It's standard CLF practise to do so isn't it ?, or are you
> > talking about something other than QUERY_STRING ?
> 
> No, I *am* talking about QUERY_STRING.  This seems kind of ..er
> *stupid*....

This is another case where there's data missing from the log: 
the actual object retreived.  Removing QUERY_STRING is easy - chop off 
everything after the ?.  Removing PATH_INFO is harder.  And finally 
determining this variant of the content-negotiated request for "mother" 
was served is impossible. (was it mother.gif? mother.jpg? mother.ps?)  

	Brian

--=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=--
brian@organic.com  brian@hyperreal.com  http://www.[hyperreal,organic].com/