We have a very huge cube almost 200m rows with exponential growth rate. I have another cube which is not so big.
I've noticed when I pivot the dimensions in the big cube, it takes more time than the latter.
Could you guys give me some advices to make this work much faster.
I've partitioned the warehouse but have'nt created any indexes so far, could this be a problem?
thanks,
Prok
Hello! Have you read the "Performance Guide" pointed to at the top of this newsgroup?
HTH
Thomas Ivarsson
|||Yeah that's definitly the best advice. The problem could be any of several.
Most likely, like you said, aggregations. 200m row cube isn't all that big as far as i'm concerned. We house a cube with ~2billion rows (180GB cube), once aggregations are designed "perfectly", you should not have an issue as long as the cube is designed well.
|||Thanks Zach84,
Yeah, I have read the pf guide before and have implemented accordingly. I must have done a bad design regarding aggregations. I'll check on that.
I appreciate your advice.
Rok.
No comments:
Post a Comment