need to decide to keep data in DSO or not

[Former Member]

Hi all,

Could someone please help me with the following scenario.

We have some data that is being loaded to DSO from there into 2 cubes. One cube will hold very detailed data (just like in DSO, which was created for faster reporting), Another Cube holds abstract level details.

we were thinking to delete data from DSO selectively as and when we upload the data to details cube, to eliminate redundancy and to save hardware resources.

So, could someone help me, what is the best practice for deleting the data from DSO. Or if you think it should not be done, then explain why its not.

any inputs will be greatly appreciated with points.

thanks

curios

[former_member619464 · Accepted solution]

hi curios

I presume that DSO is loading deltas to both the cubes. In this case Before deleting the DSO data, you have to make sure that the requests are not required for further delta extraction or for reconstruction purpose.

In future, whenever there is any change in the transformation from DSO to cube (s) - let us say - in ur present transformation, u r calculating ZAMT = ZNETAMT + ZGROSSAMT and in future u r chainging it as ZAMT = ZNETAMT + ZGROSSAMT + ZOTAMT

Once such changes r done, u ve to go for a full repair from DSO. Hence u ll need the data