>>>Can you elaborate on this or point me to a source? I am >>>working for a large company that uses Hitachi SANs and we >>>are getting HORRIBLE performance for a data warehouse (i.e. >>>15 minutes to scan a 100 MB table). Ouch. There is something much more broken than the subtleties I'm talking about. By my experience it is best to stripe volumes/filesystems across LUNs that involve multiple controllers especially if using OPEN-[39EL] -- //www.freelists.org/webpage/oracle-l