cancel
Showing results for 
Search instead for 
Did you mean: 

is HANA replacing BW completly

0 Kudos

Hi all ,

         i think this is the right place that i can ask my questions is HANA replacing BI/BW i don't know exactly but one of my friend working in SAP Labs he said with in two years HANA will occupy the BI/BW he explained me what are bad's in Bi/BW(ex: data duplication e.t.c  ) and how sap replaced with HANA also is it true and also some where i find that HADOOP vs HANA ( in db point of view only) .   is HADOOP is competitor for SAP HANA these are all going in mind till now i got only satisfied answers but i don't know when i will get exact answer.

Accepted Solutions (0)

Answers (1)

Answers (1)

werner_daehn
Active Contributor
0 Kudos

Difficult to answer as you are spanning quite a range.

Let me try this way: For reporting you have the following stack

BI tools

|

Data Management

|

DataStorage

|

Transformation

|

Source

What your friend is saying, is Source=SuiteOnHana; Transformation=CalcScenario; DataStorage=None; DataManagement=CalculationViews.

When you use BW, you would have Source=ERP; Transformation=CalcScenario/ABAP; DataStorage=Hana; DataManagement=HanaTables.

With Hana native Data Warehouse you would have Source=ERP; Transformation=CalcScenario/ETLTool; DataStorage=Hana; DataManagement=HanaTables

And with Hadoop as of today it would look like Source=ERP; Transformation=Files; DataStorage=Hadoop; DataManagement=JavaMapReduce

As you see all are very similar. You replace the one transformation language by another, store the data in Hana or Hadoop. But essentially all approaches are similar.

The most important difference is for the first, where there is no second DataStorage needed. You report directly on the source data, hence do not need a Data Warehouse Hana database, a BW Hana database, a Hadoop database.

Not having to store the data twice can be an advantage but it has sever limitations as well. For a quick "Show me the revenue of customer ABC for the last years", copying the data into a second database is nonsense, I agree. This is a report on operational data, it should be based on operational data.

But for Business Intelligence analysis like "Why am I doing so well in region A for the product category X but not in region B?" you need much more logic. Can this logic be applied at query time? Very likely not, hence you can either not run these queries, they will run for multiple minutes even on Hana, or the few changed records of yesterday are merged into the Data Warehouse database, which data model is optimized for these kinds of queries.

Some more details to be found here: