May 17, 2012 at 12:33 am
I need some suggestions from you,
We have a cube with 13 dimensions and 2 Measure Groups which have 33
partions each.
As applied PROCESS UPDATE for Dimensions and PROCESS FULL for cube
taking around 12 hrs to finish processing.
Im planning to reduce the processing time for this im planning to
apply process incremental for cube.
PROCESS UPDATE for Dimensions
PROCESS FULL for Recent Partions
PROCESS INDEX for rest of Partions.
This will surely reduce processing time but if we apply process index
then if we get any of the updated data for those transctions it wont
reflect the changes.
There is a chance of getting updates of those trancsctions with latest
extract.So it wont work
Anyone suggest me the best approach for this process.
Please let me know if you need more info.
My Lead suggest me ( write a process which finds which partion will be
impacted with the new loaded transctions and refresh only those
partions)
Ex: If Cube was refreshed on 10th may we need to check the
trancsctions loaded after that date and which partion effected those
transctions and refresh only those partions)
Is it possible to do this approach.....
May 18, 2012 at 8:44 am
Hello,
Ideally it would be beneficial to have your partitioning so that archived or static data is in a separate partition. If it is not possible to change the partitioning in this way I would agree that you will need some type of logic to determine which partition is affected by data updates.
I am assuming your server is using SSAS 2005?
James
Viewing 2 posts - 1 through 1 (of 1 total)
You must be logged in to reply to this topic. Login to reply