SAP Smart Data Integration (SDI) and difference vs Smart Data Access (SDA)

HANA as data management platform needs to connect to other data sources to get a data. There are many integration patterns to follow to enable the connection. There is an ETL processing, real time replication, data integration etc. Each of them has its own advantages as well as disadvantages. With the introduction of SAP HANA Smart Data Integration (SDI) the SAP is pursuing to get the best of some of them.

SAP Smart data integration (SDI) is an extension of SAP HANA Smart Data Access (SDA). The SDA leverages the concept of virtual tables (data federation type of thing). That means a data replication is eliminated by providing a virtual layer that abstracts the underlying data sources. Metadata of tables of the source system are imported as virtual tables into the HANA. On top of the SDA the SDI brings broader set of preinstalled adapters from Data Provisioning Agent (DP Agent) component. The adapters in case of the SDI are no longer part of HANA DB. They run as separate processes in DP Agent component.

The SDI is part of the Enterprise Information Management (EIM) solution of SAP. Its purpose is to connect data sources to the HANA, provision data from them and load the data from them to HANA.

As the SDI brings more functions into the EIM process, some customers are opting for this option. Therefore e.g. in SAP BW area they are migrating the SDA based data sources to the SDI one.

 

SDA is available since SAP HANA 1.0 SPS06

SDI is available since SAP HANA 1.0 SPS09

More information:

SDA

SDI online docu

2400022 - FAQ: SAP HANA Smart Data Integration (SDI)

2180119 - FAQ: SAP HANA Smart Data Access

2600176 – SAP HANA Smart Data Access Supported Remote Sources

SDA - Database Statistics for Virtual Tables

In order to speed up data processing for objects leveraging HANA Smart Data Access (SDA) there need to be a database statistics build. It is a concept known in many databases. Once the table is indexed, no of records is known, meta data is described etc. the processing of that table is much faster.  It is valid for tables consumed in SAP HANA or SAP BW based on SDA. As the SDA doesn’t store the data in HANA DB persistently it uses virtual tables instead also those can be indexed in order to run the SQL statements in efficient manner – so called efficient query plan.  

In SAP BW there is an ABAP program called RSSDA_CREATE_TABLE_STAT to be used to create database statistics for HANA Virtual Tables. BW’s objects like InfoProviders (aDSO and Open ODS Views) and DataSources are supported by the report. There are three types if the DB statistics that can be generated by the program:

 

1. Record Count – very basic one, providing only no of records

2. Simple - basic statistics like min, max, null count, count, and distinct count

3. Histogram – most detailed type, it estimates a data distribution in a single-column data source

 

To incorporate the DB statistics creation for SDA based data sources in BW use ABAP process type in process chains. Higher versions of BW 7.5 even have a new process type (SSDA_CREATE_TABLE_STAT under category 'Other') available for including the creation of statistics in BW Process Chains.

More information:

1990181 - BW HANA SDA: Create Database Statistics for Virtual Tables of Open ODS Views and Near-Line Storage (NLS) Archives

SAP HANA SQL and System Views Reference

SAP HANA Smart Data Access (SDA)

Within an advent of SAP HANA there was a new technology introduced to enable bringing data from external sources to the SAP HANA. It is called SAP HANA Smart Data Access (SDA). The data stored in the external databases appear if they were local tables in SAP HANA. There is no need to transfer the data physically from those external sources to the HANA in order to use them.

It uses a concept of virtual tables which point (via mapping) to remote tables in different data sources. In the HANA an SQL statements can be executed to process those data. The HANA query processor process these SQL statements and execute part of them against target database returning results to the HANA.

There is a number of the external databases that are supported. Depending on their hardware platform it varies from: SAP’s DB portfolio (HANA DB, IQ, Adaptive Server Enterprise (ASE), Event Stream Processor, HANA Smart Data Streaming, MaxDB), MS SQL Server, Oracle, Apache Hadoop, Apache Spark, IBM Netezza, IBM DB, Teradata, etc. The SDA was introduced in version of SAP HANA 1.0 SPS06 (in 2013).

 

More information:

1868209 - SAP HANA Smart Data Access 1.0 Master Release Note

2352696 - SAP HANA Smart Data Access 2.0 Master Release Note

Online docu SAP HANA Smart Data Access