Re: Raw vs. cooked

FProse wrote: > > Using AIX 4.1 and DynamicServer 7.2 on a four processor RS6000, I > found cooked beat raw by a margin of 15%. This was running against > Clarion array disks, the raw was non-raid, the cooked was raid-5. > > The local Informix people checked my results and went away shaking > their heads. You are comparing apples and grapefruits dude! You just proved my point! If your RAID set was made up of 5 drives it should have been 4 times faster than the single RAW drive, but the 50% RAID5 write penalty brings that down to it should be 2X faster. However, you tested it out at only 15% faster which works out to being 60% slower (off the top of my head you are free to do the math but it's close) than a RAW RAID5 disk farm would have been! I have tested RAID5 RAW -vs- RAID5 COOKED, RAID1 RAW -vs- COOKED, and RAID10 RAW -vs- COOKED and the results are always the same. RAW is 35% faster than COOKED files and 15% faster than COOKED devices! You know what? My American made Torus station wagon is faster than my daughter's Japanese made bicycle - proving conclusively that Japanese vehicles are slower than American vehicles! Oh damn, two flames and it's only May. Now look what you made me do! Art S. Kagel