No the data was partitioned along time with each period living in it's own fact table. The data was then built into cubes along our partitioning strategy. I don't know I would ever design something with billions of rows in a single table. If I had knowledge of the expected data size at design time I would put serious consideration into partitioning strategies for the data.
Yes, we will be partitioning the data along monthly processing date. But the over all table will have billions of rows and each partition will have 100 to 200 hundred milliion rows per month. I am wondering about a table and not each partition.
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.