Could anybody give me a suggestion for the following issue?

We have a master DB to store all of our clients Data and we use AS to
generate reports. On the client-side, each client only can see its own data.
So we meet an issue regarding how to design the cubes and dimensions.

We already have two choices:

1. Creating dimensions against the master DB and creating Cubes with
partitions to seperate clients' Data. But this way has two problems -- when
show the cubes in OWC10.Pivottable, all clients' data will be exposed from
dimensions to every client. Moreover, how to create the local cube from each
partition?

2. Creating a set of sub-DBs from the master DB to seperate each client's
data. Then create cubes against sub-DBs. But this way is hard to maintain if
any changes in Cubes and Dimensions. have to change for each sub-DB.

Has anybody met a similar situation?


Yan

Hi Yan, interesting challenge, indeed.

My suggestion would be a "hybrid" between your two suggestions.

I would have one OLAP DB where all dimensions that can be shared, are
shared, so that maintenace is easier. Then, I'd seperate the customers into
different physical cubes. If any information is necessary and private just
for that customer, I'd put that information in private dimensions (not
shared) for just the customer's cube. This way, the information is seperated
by physical cubes, but the shared dimensions can all be maintained in one
OLAP DB.

Good luck.

- Phil


"yan" wrote:

Hi Phil

Good Point! Although we have few shared dimensions left. But this way is
still better than creating sub-DBs.

Thank you very much.

If you have any other ideas, please let me know.

Yan


"SQL McOLAP" wrote: