problem with change package source

[Behnamqmarsi]

Hello to all friends

I have two basic cube(first cubes , on data relating to previous years and the second Cube is related to recent months) and  I've created a virtual cube . And numerous reports have made on the Virtual Cube.

After a one year needs to be considered is the second cube in a virtual  Cube. In this case, the new virtual Cube name change and therefore the reports will not run properly

Please advice in this problem To previous reports run without error

cubes type is dynamic cube and design with cognos cube designer 10.2.2
thanks

[MFGF]

Hello to all friends

I have two basic cube(first cubes , on data relating to previous years and the second Cube is related to recent months) and  I've created a virtual cube . And numerous reports have made on the Virtual Cube.

After a one year needs to be considered is the second cube in a virtual  Cube. In this case, the new virtual Cube name change and therefore the reports will not run properly

Please advice in this problem To previous reports run without error

cubes type is dynamic cube and design with cognos cube designer 10.2.2
thanks

Hi,

This seems to be a fundamental flaw in the design of your strategy. No two cubes will contain members using the same Member Unique Names (MUNs), therefore any reports written against one cube that reference its members will not work if you switch the cube to a different one.

Why are you adopting a strategy to use different cubes for different years?

MF.

[Behnamqmarsi]

HI

Yes. All cubes are quite similar and are different in  time periods

Because of the long duration of filling defined in memory aggregate and Issues such as refresh or update,we  have to build virtual cubes for small periods of time, Thus, when creating new virtual cube, made previous reports will not run


Please help me on this Issue :'(

[MFGF]

HI

Yes. All cubes are quite similar and are different in  time periods

Because of the long duration of filling defined in memory aggregate and Issues such as refresh or update,we  have to build virtual cubes for small periods of time, Thus, when creating new virtual cube, made previous reports will not run


Please help me on this Issue :'(

Hi,

I thought I was already trying to help?

You're making a statement here that sounds reasonable, but isn't viable. Building virtual cubes for small periods of time and swapping those cubes out isn't the one and only solution to your in-memory aggregates taking a long time to build. Have you looked into using in-database aggregate tables? These can drastically reduce the time taken to populate your in-memory aggregates, and might also reduce the need for some of them. I'd further assume that the vast majority of data you want to use is static, and only a very small subset of recent data is changing? Have you thought of using a similar strategy where the static data is in one cube (using caching and in-database/in-memory aggregates) and the recent data is in a small uncached cube with no aggregates, then having these merged in a virtual cube? As the small cube is uncached, this might reduce the need to keep re-starting and re-populating your aggregates?

Have you read the Dynamic Cubes redbook? It has lots of good advice to give on the design and use of Dynamic Cubes, and might give you some further ideas on what is the best strategy for you to adopt here.

Cheers!

MF.