httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Andrew Wilson <>
Subject Re: custom logging again++
Date Thu, 23 Mar 1995 23:46:50 GMT
Rob H, on the vexed question of having many log files:
> I like frosting. So while we wait for the next big feature upgrade,
> do 90% of the team sit around doing nothing ? Rob T can continue
> his good work on content-negotiation, but it's pointless having the
> rest of us do nothing.
> Anyway, there's no harm in talking about wish lists items.

D00d.  Somewhere down the line someone's going to have to get all those log
files together and try to figure out who to send bills to for using the
services people provide over the WWW.  If log files contain some idea of who
was involved, ie, more than just an IP address (which is next to useless for
billing models, cuz 90% of people have roving IP addresses curtesey of their
service providors anyway!), then it'd be more of a help.  Hiving off hits
against a certain portion of your web space into some special logfile is fine
in principal, but do you really want the server to be making your end-of-day
stats analysis easier by chewing on the CPU at access time?

By all means record useful info in log files, but don't go wasting time
putting pseudo analytical hacks into the server, 'cuz a standalone PERL script
is going to do the job much better anyway.

I have spoken,

     Andrew Wilson	     URL:
Elsevier Science, Oxford   Office: +44 0865 843155     Mobile: +44 0589 616144

ps.	When we've got all this performance and configurablity nonsense out of
	the way [ ;) ] we can get down to the real meat.  Security, and useful

View raw message