Completely depends on your situation -- big files, small files? Lots of writes or reads or both? Growing tables/files or lots of large dropped tables?
If you're into small files, go Reiser. Big data, JFS or XFS. EXT3 is slow, but, IIRC, it also is a true data journaling filesystem while some of the others aren't (and that's as deep as I go without a life-vest, if you know what I mean!).
Based on what I've seen so far, JFS and XFS do not yet have a solid track record of stability with MySQL. This does not mean they could not be good - I just do not trust them yet. I do vaguely remember a support case when a very strange corruption happened on either one of them - now cannot recall which one. In 3 years of handling MySQL supoprt (2000-2003), I do not recall reports of table corruption on a ReiserFS file system. I do remember corruption cases in the early ext3.
If you are using InnoDB, the choice of the filesystem should not be that critical because of full caching. For MyISAM, having a good file system is very important - unlike InnoDB, MyISAM does not cache the data itself, and relies on the OS cache. So you do have a lot of read/write syscalls. To illustrate the difference - I recall a case when performance on MyISAM was terrible over NFS (well, that is to be expected), but once the table was changed to InnoDB, it improved drastically.
-- Sasha Pachev Create online surveys at http://www.surveyz.com/
-- MySQL General Mailing List For list archives: http://lists.mysql.com/mysql To unsubscribe: http://lists.mysql.com/[EMAIL PROTECTED]
