Hi all,
Just to keep you posted: I've begun doing some serious refactoring of
the server code, while working on the preforking patch. I've also
updated the patch to the 2.x codebase.
I've implemented a rudimentary locking mechanism in the scoreboard
accessors. In fact, so simple it must be prone to races.
I've only tested it in a simulated 2.x environment: netcat rulez. The
number of processes scales real nice. I've also backported to 1.2.3 and
its running smoothly on my production server.
http://sourceforge.net/tracker/index.php?func=detail&aid=831675&group_id=85894&atid=577644
--
________________________________________________________________
Paul Stevens paul@nfg.nl
NET FACILITIES GROUP GPG/PGP: 1024D/11F8CD31
The Netherlands_______________________________________www.nfg.nl
Just to keep you posted: I've begun doing some serious refactoring of
the server code, while working on the preforking patch. I've also
updated the patch to the 2.x codebase.
I've implemented a rudimentary locking mechanism in the scoreboard
accessors. In fact, so simple it must be prone to races.
I've only tested it in a simulated 2.x environment: netcat rulez. The
number of processes scales real nice. I've also backported to 1.2.3 and
its running smoothly on my production server.
http://sourceforge.net/tracker/index.php?func=detail&aid=831675&group_id=85894&atid=577644
--
________________________________________________________________
Paul Stevens paul@nfg.nl
NET FACILITIES GROUP GPG/PGP: 1024D/11F8CD31
The Netherlands_______________________________________www.nfg.nl