Re: Raw vs. cooked

You can do your math any way you want to -- I do mine with a stop watch!. The bottom line was the cooked ran faster on a 1+ hour benchmark. BTW, it was raw disks, 5 db spaces, one per spindle.

A lot of the performance advantage of cooked compared to raw has to do with UNIX filesystem buffering/engine buffering. If the engine's bufferpool space is small compared to the total real memory space, then I would expect that cooked filesystem would significatly outpreform raw. If the informix bufferpool was large compared to total real memory space, then I would expect raw to perform better, simply because the UNIX filesystem buffering would require paging to become active. FProse wrote: > You can do your math any way you want to -- I do mine with a stop > watch!. The bottom line was the cooked ran faster on a 1+ hour > benchmark. > > BTW, it was raw disks, 5 db spaces, one per spindle.