SLT vs ODP/ODQ for extraction to SAP BWonHANA
SLT for BW
SAP LT Replication Server
offers a useful alternative for data transfer to BW in the following cases:
• The tables
in the source are simple tables without joins or transformation logic.
• The
DataSources (extractors) that you want to replace using the SAP LT Replication
Server are DataSources on simple tables/views that do not provide a delta
mechanism and only contain minimal extractor logic.
Two interfaces
are available for transferring data using SAP LT Replication Server:
Operational Data Provisioning
•
Data transfer using operational data provisioning is supported for tables from
SAP systems only.
• SAP
LT Replication Server provides the operational data provisioning infrastructure
with the source tables as delta queues. The data from the delta queue can be
replicated in BW as a subscriber.
• If
you use operational data provisioning, you can load the data directly into the
InfoProviders (bypassing the PSA layer) by using a data transfer process. The
ODP infrastructure (with delta queues) takes over important services such as
monitoring data requests. In addition, the ODP infrastructure has been prepared
to support BW and other subscribers with SLT data transfer (for example, SAP Data Services).
Web service
•
Data transfer using the Web service interface is supported for tables from SAP
systems and for tables from non-SAP systems.
• SAP
LT Replication Server replicates the data in a Web service DataSource of the BW
Persistent Staging Area, where the data is available for further processing.
• If
you use the Web service interface, the data can be pushed at regular time
intervals to BW, where it can be updated using real-time data acquisition.
ODP/ODQ
ODP(Operational Data provisioning) is a Netweaver based framework, ODQ(Operational Delta queue) is the queue for data extraction where
ODP is installed as a source.
ODP is a unified
infrastructure for data provisioning and consumption which supports scenarios
like:
Option
1: ODP based Data Provisioning Aspects for SAP ERP Sources
ODP allows to
skip the PSA layer and load directly with DTP from the source system into a
DSO. So this ODQ acts as replacement of BW Service API Delta Queue (RSA7) w/o
the need of collective jobs run needed earlier.
Option
2: SLT/ODP based real-time replication
SLT replication
server allows loading and replicating data in real time from SAP and non-SAP
source systems into SAP HANA environment. SLT allows real time and scheduled
data replication, replicating only relevant data into HANA. SLT Server can act
as a provider for the Operational Data Provisioning Framework (ODP) and stores
data from connected SAP systems in this framework in an Operational Delta Queue
(ODQ).
Option
3: ODP based data transfer between BW systems
ODP extractors
can be used to transfer data between BW systems.
Good
business case on Data extraction and Modelling
Re-iterating from my earlier email on this topic(attached), below is a good
example which shows how SLT replicates the data to HANA DB from ECC system and
then replicated data is modelled in HANA and accessed finally to BW using ODP.
This is a very good example for faster accessibility to data using SLT and
HANA.
Advantages of SLT
• ODP w/o SLT is
only possible for BW datasources, ODQ is a replacement of RSA7 delta queue. So
this becomes useful when you are dealing only with SAP ERP datasources and
there is no requirement of HANA Native modelling further for Analytics.
• With SLT, its
possible to replicate source tables from SAP as well as non-SAP systems which
can further be used in native HANA modelling. Additionally, it also supports BW
modelling. Here is the advantage of SLT which offers more flexible options.
• So, SLT is most
useful when you have multiple SAP and non-SAP sources of data want to
leverage the HANA Native modelling aspects and this is more aligned to future
SAP roadmap.
Advantages of ODP/ODQ without SLT
- There is no additional licensing and hardware cost
required for ODP extraction without SLT. Hence no additional maintenance
effort is required here.
The selection should entirely be made based upon the business
requirements and needs.
