Extended Analytics / Cube Designer - Transport

[deepesh_kumar]

Hi All,

Tools - BFC, Cube designer / Extended Analytics

Landscape - Development - Quality - Production

Versions - FC 10.0 and Cube designer 10.0

Question -

Do we transport the objects related to Cube Designer / Extended Analytics or do we create directly in each system.

The objects in Cube designer that I can think of for transport are - Solution, View and Deployment. How ever in the tool I can see only option for Export and Import Deployment.

How is it handled normally? Can anyone share the experience.

Regards

[Marc_Kuipers · Accepted solution]

Hello Deepesh,

There is no 'transport' option as such for these objects. Technically, all these (solutions, views and deployments) are stored in the BFC database and therefore linked to meta-data in this DB

If you want to move a view/solution, you should export it and then re-import it

Then, make sure to change the deployment to the correct MS Analysis server and repository.

Thanks,

Marc

[deepesh_kumar · Accepted solution]

Hi Marc,

Thanks a lot.

In my case, using Start Schema Deployment (Not SSAS or BW deployments). I hope what you said is applicable here also.

Few more questions on Cube Designer -

1. As per my understanding if we add any new dimension member in BFC (say a new account) it will get reflected in the cube designer existing Views with out any changes. Please confirm.

2. If a new dimension is added in the BFC (New Analysis dimension), do we need to create a new View? or will it work with the existing Views

3. If a new property or characteristics is added in the BFC existing dimension, is there any adjustments to be done in Cube Designer so that it will get reflected in Cube Designer?

Thanks in advance.

Regards

Deepesh Kumar

[Marc_Kuipers · Accepted solution]

Hi Deepesh,

The same applies to Star Schemas, the main difference being that the star schema itself (the fact and dimension tables) are also stored in the relational DB.

1. Yes, the member will be present, but I believe it will appear as unselected (unless of course you have selected your members by some filter or characteristic and this new value is within that selection)

2. This is a tricky one.

If you introduce a new dimension (which can only be a user-defined dimension and therefore you need dimensional analysis) it depends what was done with 'old' data.

In Designer, the new dimension will be visible, but if you are using multi measures, you would need to select this new dimension in the "Dimensional Analysis" tab.

You would now be in a situation where you have a new measure and you would as a first step you would need to deploy. Now, if in BFC you introduced this new dimension in (for example) 2014.04, all your data prior to this period would not have any values for this DA.

In the cube this means that if you selected this new measure, you would not see any data for 2014.03, 2014.02 etc.

This means the end-user needs to select both measures in the report- and this can be quite confusing. You can use 'single measure' mode to avoid this

3. If you want to use this new characteristic as a hierarchy, you would need to select it in the hierarchy tab and re-deploy

Thanks

Marc

[deepesh_kumar · Accepted solution]

Thanks Marc,

Answers from you help to close open points confidently .

Regards

Deepesh Kumar