[8533] in www-talk@info.cern.ch
Re: What's next after the common log format?
daemon@ATHENA.MIT.EDU (Darren New)
Thu Jun 29 09:55:49 1995
Date: Thu, 29 Jun 1995 09:44:44 +0100
From: Darren New <dnew@sgf.fv.com>
To: Andrew Payne <payne@openmarket.com>
Cc: Brian Behlendorf <brian@organic.com>, www-talk@www10.w3.org
In-Reply-To: <199506290116.VAA27121@aegean.openmarket.com>
> We've actually started work in the other direction, where the server might
> accumulate the logging information for some user-specified period of time
> (several seconds, typically), and dump the information in a single I/O
> operation. You can still do all of the fancy stuff about displaying
> response times, etc., but they'll be a few seconds behind the curve.
Actually, I just finished code to do such a thing as part of a
microtransaction accumulator. It maintains numerous small databases and
logfiles in memory, dumping them out periodically, talking to all the web
processes (or whatever) over sockets. (More efficient and reliable that
trying to make file locking work properly, plus it works with replicated
servers.)
It's not hard... it's essentially what Syslog does. If you worry more
about the API than the file formats (or at least divorce the two) you'll
be much better off when you want to replicate web servers or such stuff.
--Darren
