2 .ETL :
Developing a link between source to target
- Data Source :
This is a structure, which is created in the source system and replicated to the BW system. Maintaining a data source in a source system is doing any changes related to the data source. It could be adding/deleting fields or checking the flags in the data source like the hide and selection flag. Also, if the datasource is provided in the LO Cockpit, changes can be made there too depending on how the extraction of data is made for that data source. Extraction could be of flat file extraction or of multiple flat structures ie., hierarchies.
What data has to be extracted from source
- Persistant Staging Area (PSA) [2DM table]:
PSA is a automatically generated by system itself when we create data source.In generally we will be cleaning these PSA tables on regular intervals of time .
- Transformations :
This is to transform data from staging tables and moved to targets.Some times we will write some ABAP code to transform the data , This is called routines.
3. Reporting
Reporting tool may be different from client to client , default tool provided by SAP is Business Explorer (Bex) . Sap has recently acquired BO ( Business objects) , Soon this will replace Bex.
Combination of 4 applications :
Using analyser we can report in the form of excel sheet
Input for Analyser is Query & info Provider
Using this we can design the query according to requirements of end user and we can execute this in browser
Input for Query designer is Info Providers
We can format the query output using report designer .
Input for report designer is Query
Using this we can see query output in our own designed webpage
Step 1 ) We can design webpage .If you know HTML you can design your own webpage ,else we would be having drag and drop options available to create webpage .
Step 2 ) Data Binding --Assign input for the web page
Query
Info Provider
Query View (When we design small reports using analyser , that will be called as Query view)
Suppose if we want to see two table output , This can't be done using query designer .We need to use WAD to design such requirements.
4) Performance Tuning
- AGGREGATION : This is hard to understand in the beginning , Aggregates are like a smaller cubes on the main cube to improve query performance
- COMPRESSION:
- PARTITIONS
- INDEX
- LINE ITEM DIMENSIONS
- OLAP CACHE MEMORY
- BI ACCELERATOR INDEX
5)Broad Casting
Sending the reports to concerned end users
- When to Send ?(Time)
- How to Send ? (Format)
- Where to Send ?(Location)
No comments:
Post a Comment