cooked vs raw disk in an Data Warehousing environment

The enviroment is HP N Class machines running HP 10.x. Informix 9.1 Does it really make a difference in having raw disk for storing data in a DSS environment, as compared to cooked file system. There will be a monthly incremental load to the Data Warehouse and rest of the time there will be only queries. The database itself will not have transactions. While we are at it, is it easy to manage dbspaces and fragments with cooked file system as compared to raw disk. What about adding extents or dbspaces. Is cooked file system easier to handle. thanks and regards. Sent via Deja.com http://www.deja.com/ Before you buy.

(I thought you had to run HP-UX 11.0 on N classes. What do I know?!) It does make a significant difference on HP-UX. A little empirical evidence I have on IDS7.31/HP-UX 10.20 is around 15% improvement in large insert times. Raw disk is quite easy to administer, particularly when using an lvm like, er lvm! I can see no reason to use cooked disk unless you're an absolute novice, or your SysAdmin is a bozo. <rkinformix@yahoo.com> wrote in message news:8rvlhc$fh7$1@nnrp1.deja.com... > The enviroment is HP N Class machines running HP 10.x. > Informix 9.1 > > Does it really make a difference in having raw disk for storing data in > a DSS environment, as compared to cooked file system. There will be a > monthly incremental load to the Data Warehouse and rest of the time > there will be only queries. The database itself will not have > transactions. > > While we are at it, is it easy to manage dbspaces and fragments with > cooked file system as compared to raw disk. What about adding extents > or dbspaces. Is cooked file system easier to handle. > > thanks and regards. > > > Sent via Deja.com http://www.deja.com/ > Before you buy.