In Oracle hash paritioning do we have control over hash value based on which partitioning is done? My issue is -
We have to make a design for datawarehouse application and in the landing area we are planning for range-hash partitionining based on the keys (and their volumes). Now, if we use range-hash paritioning, if the hash values of those keys which has the large number of records turn out to be the same. then I may face problem, becuase one partition will have large number of records, whereas the other partitions would be hardly occupied (wasting resources). Currently we do not have the volume statistics per key values, so do we have control over the hash partition values so that I can manually define partition based on hash value?
As an alternate approach, can we do paritioning dynamically based on the volumes when the data is loaded in the landing area?
October 11, 2005 1:06 AM
October 14, 2005 3:35 AM