I have to disagreeMarch 19 2012 at 1:32 PM
No score for this post
|Cameron Lackpour |
from IP address 126.96.36.199
Response to Tim - you are correct for BSO but not ASO
Are you stating that BSO is *not* disk bound? In my experience, BSO is disk bound as well.
Let's face it -- if Essbase has to go to disk for *anything*, it's slow. There are plenty (and by plenty I mean most) of BSO Essbase databases that *cannot* fit in RAM. Can. Not. Fit. And they are correspondingly slow.
It has been quite a while since server CPUs were slow enough that fast drives could overwhelm them with data.
Now *ASO* databases fitting in RAM -- and then being CPU bound, well, that I can understand.
Of course everything I wrote about BSO changes with (I think I am getting this right) the Times Ten in-memory functionality that's in Exalytics.
- I think our overlapping responses are getting confused - DanP on Mar 19, 2:54 PM
- Re: I think our overlapping responses are getting confused - Tim Faitsch on Mar 19, 3:52 PM
- Agree on ASO Cache Size but that was not the question - DanP on Mar 19, 5:46 PM
- Re: Agree on ASO Cache Size but that was not the question - Tim Faitsch on Mar 20, 7:37 AM