*** i'm writing a file cache to use together with a web application. pretty straight forward: when a file first is requested, load it from file and put it into memory. from then on use the copy from memory. the problem that i have is that there is NO performance gain at all when reading a file from the cache as opposed to reading it from the file system. my programming may not be perfect but i should see some gain. it looks like linux itself is caching the file, because the first access from the file system is a little slower than subsequent access from the file system. is this possible? can i turn the file caching on the operating system level off (for testing purposes)? any other ideas? oh yeah, is there source code for a file cache available somewhere? thanks, - martin / martin gloeckle / senior software engineer / [EMAIL PROTECTED] / / think new ideas, san francisco / http://www.thinkinc.com / / san francisco - new york - los angeles - boston - atlanta - seattle - london - sofia / ---------------------------------------------------------------------- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]