Hi Joe,
I agree with Immy it depends on your scenario or business needs.
But i still see some other underlying questions in your primary question.
What is your designed landscape.
1 Do you have a Data Mart as a subset of the data warehouse and do you use
ROLAP.
2 Or is your OLAP cube a MOLAP and is this also what you call a Data Mart.
3 Or is your Data Mart a pre-phase for the data warehouse and you use either
MOLAP or ROLAP on the Data Mart.
4 Or do you use the term Data Mart, for a small standalone Data Warehouse.
To me these are very interesting questions, because there can be a lot of
misunderstanding over a simple terminology as a data mart. Cognos for
instance sees an OLAP-cube as a data mart and with this a data mart as a
subset of the data warehouse. In some Microsoft outings i see the same
descriptions, but they are not consistent.
Okay, what has this got to do with your question?
In scenario 1 it is obvious. There is no need for a upload if the complete
data warehous is only uploaded once a month.
In scenario 2 it is the same, if the data warehouse is uploaded once a month
there is no need to upload the MOLAP cube more frequently.
In scenario 3 there is a real struggle. Because you can upload the
fact-table but what about the dimensions. Are the dimensions only used inside
the data mart or across data marts. Only for these cubes or across cubes.
I suppose that in this case your scenario is scenario 4.
Well i don't know if this was helpful or if it was just a way of getting my
frustration over the usage of the term Data Mart of my chest
;-0
--
Johan Blad
Business Intelligence consultant
Ultimate Software
[quoted text, click to view] "Joe" wrote:
> We have an existing data mart where it's load cycle is the first Saturday
> of every month. Besides making the data available quicker, are there
> more pro's to shortening the load cycle to every day or once a week? It's
> a very small data mart with most users (under 10) accessing it via cubes.
