> Could you not create a raw device and install a filesystem on it for > your data? Access to the raw device would bypass the Linux cache.
Your application would need to be specifically coded for that kind of access, and know how to manage raw blocks. Dave: you can't turn the Linux cache completely off, but may be able to accomplish what you want by manipulating kernel parms on the commit rate and frequency of scans. There is a paper on tuning that by a couple of guys at RH on what kernel vars to tinker with. I think it was mentioned on this list a while back by Alan Cox. Contact me off line if you can't find a copy. -- db ---------------------------------------------------------------------- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390
