Desicion Stream Catalog

[swarnapuri]

Hello All,

I am new to Cognos Development environment. As of now, we are using MS Access for storing Decision Stream catalog information. Is there advantages and disadvantages for using Oracle for storing the catalog?

Also, Is there a place where I can find best practices for using Decision Stream/Impromtu/Power Play. If someone in this forum has a similar kind of document, can you please share with me?

Regards,
Vinod Swarnapuri

 

[Kaiaualad]

Some best practices for decisionStream are:
-Create standardised naming conventions for columns and tables and catalogues.e.g D_dimension Table or F_Fact table
- If using more than one catalogue, set environmental variables using batch scripts which centralises them.
- Create a Dev and Production environments preferably on separate boxes but have identical directory structures so pregression through the envirnments is easier.
- Utilise lookups or dimension references in builds instead of using the SQL in the datastreams - its more efficient.
- Standardise the use of common functions in DS catalogues
- If extracting source data ,extract all fields and load in source_database, so builds can hit that and not the transactional databases.
- Utilise a database_alias text file for database connections to hold all connections in one file for ease of changes.
- Drop all staging tables at the end of a build.
- create a rotate function that builds to a temp table, then copies it over the existing table.This ensures if there is a table lock on the existing table the build wont completely fail.
- build in a reconciliation build to match count rows or sum data between source and stage, and source and target.
- test every stage of development, UAT and production regression testing.

Good luck