analyse this linux file sharing related question 3516

New content (from an SQL database) can be request by user for one of the thousands of "id's" , so there is no question of precreating the presentation layer (html) as not all precreated content would have a surity of access. If we did that then it would be a waste of cpu cycles.

Due to heavy amount of content which belongs to various "ids" and various "cats" and which changes in real time (in an SQL database), my scheme is as below:

user requests: id4cat3.htm (one of the thousand id and cat combinations)

if exists apache serves this file not requiring any script runs (this saves cpu cycles)

otherwise apache calls index.php?id4cat3 index.php is a script that deciphers id4cat3 as id=4&cat=3 parameters then index.php presents latest html content for id=4 and cat=3. then dumps the output to id4cat3.htm file (so any future users requesting id4cat3.htm will get it off the shelf, saving script runs)

id4cat3.htm is deleted by cronjob every 10 minutes (to allow index.php to re-create id4cat3.htm when it is requested again by a future user, presenting some new content that "just arrived" for id=4 and cat=3)

The process repeats.

Mike