This content has been marked as final. Show 6 replies
This link may help you...
Partial clear takes more time than full clear.
You do not have to clear agg views but Partial Clear takes time if aggregate views are present.
You can try slices but I am not sure if it works that nicely performance wise specially when agg views are present.
Query hints are used for building aggregate views.
You can refer to dbag to get more idea about loading to slices and Query hints.
Thanks for the reply..
my main motive is to reload same refreshed data for same intersection with less downtime..
if i will clear aggregation then do partial clear followed by data load and aggregation then it will be take more time(not sure ..i am assuming it as aggregation take more time)
just want to clear data set with aggregated view and load data to that intersection and aggregate that slice...
looking for best approach or if any one as actually did tried that
you should note the syntax you were givven was for a logical clear wich is faster but creates slices. If you do a physical clear it will take longer
Lets say Jan->FY12->Actual->ITdepartment needs to cleared and reloaded with new set of data in cube which has aggregated view.
i am thinking this can be done without clearing aggregation but i dont know if i can reaggregate the new slice of data
step 1) physical partial clear without clearing aggregation
step 2) reload new refresh data for same intersection with aggregation
step 3) how to reaggregate the new slice of data?
is it the right approach ?
will this be faster than clearing aggregation and then loading the slice of data?
If you do not clear the aggregations, the slice will get its own aggregations mimicing the aggregations on the main cube