Hi All,
We had used ABAP extraction to extract mass data from BW in DS 3.1
Q(1)
Is this still valid in DS 4.0?
Q(2)
We came to know of a method using Open Hub from this link too.
https://wiki.sdn.sap.com/wiki/display/BOBJ/Reading+SAP+BW
Has this got any advantage over the ABAP extraction method?
ganeshm (BOB member since 2011-03-17)
In my experience, using Open Hubs for data extraction from BW has a couple of disadvantages. The first is that it is significantly slower than ABAP extraction. The second is that Open Hubs are not ideal for large volumes of data. If you only need to load 500k records this won’t be a problem, but if you need to load 100MM records it will be. The third and last is that it is necessary to customize BW in order to be able to use Open Hubs. More specifically, you need at least an Open Hub table, a process chain, a DTP and a transformation.
However, Open Hubs offer one big advantage over ABAP dataflows: delta support. If you need this, Open Hubs are probably the way to go.
weeksi (BOB member since 2012-02-20)
One additional thing to consider, if you read from cubes, understanding all the table might be too difficult and hence you use ODP?
Werner Daehn 
(BOB member since 2004-12-17)
Hi Werner, is SAP going to do anything about low performance of OpenHub-based load?
As for extraction from cubes, it seems like the better option (because OHD is ungodly slow) would be to shove the data into an intermediary table/DSO and then fetch it via ABAP flow.
Codehopper (BOB member since 2010-05-14)