on 09-10-2009 8:12 AM
Hi All,
I have 5 years of historical data that needs to be loaded.
From data modelling point of view, is it better to create a few cube for it, let's say one cube per year or one cube per company code, or shoud I just use one cube and then have it partitioned ?
Please advice. Thanks.
Thanks Ralf.
One year data could consists of more than 2 million of records.
If we separate one cube per year, will this mean that every year we have to create a new cube and change the loading destination ?
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Hello,
it depends a bit on the usage and size of the cube, e.g. are most reports only reding data from the current year, how much data i expectedper year etc.
Nevertheless, I would recommend to use both! Use 1 cube per year and connect them with a multiprovider. For enhanced reporting performance the multiporvider hint (SAP note 911939)can be used.
Then the cubes can be partitioned by an additional time characteristic, like 0calmonth or 0fiscper.
Best regards,
Ralf
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
User | Count |
---|---|
90 | |
10 | |
10 | |
10 | |
7 | |
7 | |
6 | |
5 | |
4 | |
3 |
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.