Best practice for managing multiple schemas

[patrickbachmann]

Hi folks,

I'm working on pushing data into HANA from a BW Open Hub using BODS.  I am next going to model some views over these tables and consume them using third party Micro Strategy.  So in our original test we have setup a BODS user in HANA that created the tables and loaded them directly into the private schema for the BODS user (NOTE: All other SAP replicated data sits in a separate default schema.)  My question is what are some of your recommendations for handling these schemas?

For example;

1) Keep default schema with only SAP replicated tables, and the new BODS schema with only the pushed BW data and allow a single microstrategy system user to access only this new BODS schema

2) Instead of pushing BW data to BODS Schema, push data directly into the default schema so that all this new data is in the same place as replicated SAP data and then allow microstrategy system user to access the default schema

3) Some other option?

I'm just wondering if there's some advantage to either method and am trying to think of future ease of use/best practices for securing all of this.

Thanks,

-Patrick

[former_member184768 · Accepted solution]

Hi Patrick,

Personally I will be interested in understanding why you choose to have MSTR as the front end tool. Even we have the same for some other reasons.

The approach we are going ahead with is as follows:

Publish the DSOs to HANA as Analytic views and consume the Analytic / Calc views (built on DSO generated Analytic views) through MSTR.

This way, there is no data redundancy and the BW application logic and data remains within the control of BW. The data can also be made available for reporting using BW / BO tools (if required) and no data reconciliation between multiple schemas is required. This would ensure "single version of truth" across all reporting tools.

Regards,

Ravi

[patrickbachmann · Accepted solution]

Hi Ravi, thanks for your reply.  Just to clarify though, we are not currently using BW on HANA. We have traditional BW and that's why we are using BODS to push the data to HANA.  But my question is really around the schema naming and security to access the schema.  ie: Should we use the default schema or create a second schema specific for just MSTR related items.

[former_member184768 · Accepted solution]

Hi Patrick,

My bad.. I didn't understand it quite well.

In my opinion, the data can also reside in the same default schema. I hope MSTR will consume the data through HANA Information models which will be residing in packages. The authorization for the MSTR application can be provided at the package level.

Even if the BODS data for MSTR and data in default schema reside in separate schemas the HANA Information models can still access data from both the schemas (which may lead to prefixing the schema name to the objects), unless you have same table name with different data in both the schemas.

You may choose to name the tables which store the data prefixed with BD_<table_name> to visually identify the BODS related tables with other tables, but from the access control perspective, I think it should also be defined at the HANA Information package level.

Regards,

Ravi