DATA Provisioning & Replication in SAP HANA
Mason Garcia
⚡ Smart Summary
Data Provisioning in SAP HANA loads data from source systems before any front end tool can reach it. Extract, transform, and load form the process, delivered through built-in options or external tools such as SLT, Data Services, and DXC.
What is DATA Provisioning?
DATA Provisioning is a process of creating, preparing, and enabling a network to provide data to its user. Data needs to be loaded to SAP HANA before data reaches to the user via a front-end tool.
All these processes are referred as ETL (Extract, Transform, and Load), and detail is as below-
- Extract – This is the first and sometimes most difficult part of ETL, in which data are extracted from the different source system.
- Transform – In the Transformation Part, series of rules or functions is defined for the data extracted from the source system, for loading data into the target system.
- Load – The load phase loads the data in the target system.
Because SAP HANA holds data in memory and is frequently used for real time reporting, the choice of provisioning method has a direct effect on how current the data in a report actually is. That choice is examined after the available tools are listed.
Replication in SAP HANA
SAP HANA supports two type of Provisioning tools –
- SAP HANA Built-In Provisioning Tool
- Flat File
- Smart Data Streaming
- Smart Data Access (SDA)
- Enterprise Information Management (EIM)
- Remote data
- External tool supported by SAP HANA
- SAP Landscape Transformation
- SAP Business Objects Data Services
- SAP Direct Extractor Connection
- Sybase Replication Server
At present, there are main methods of data provisioning for SAP HANA, these are –
| Methods of Data Provisioning | Description |
|---|---|
| SLT | SLT (“SAP Landscape Transformation Replication Server”) running on the SAP Net Weaver Platform. SLT is an ideal solution for Real-Time and Schedule time replication for SAP and non-SAP source system. |
| SAP DATA Services | SAP DATA Services is a platform for designing of ETL processes with a graphical user interface. |
| DXC | DXC stands for Direct Extractor Connect is a batch driven ETL tool. |
| Flat File Upload | This option used to Upload data (.csv, .xls, .xlsx) to SAP HANA. |
SAP HANA SLT Road Map
SAP HANA SLT Road Map is as below – DATA Provisioning through SLT require RFC/DB connection to SAP/Non-SAP Source System and a DB connection for SAP HANA database. On SAP SLT server, we define Mapping and Transformation. Below is a roadmap for data provisioning through SLT.
Comparing SLT, Data Services, DXC and Flat File
The four main methods are not interchangeable. Each was designed for a different latency requirement and a different kind of source, which the table makes clear.
| Parameter | SLT | SAP Data Services | DXC | Flat File |
|---|---|---|---|---|
| Latency | Real time or scheduled | Batch, scheduled | Batch, scheduled | Manual, one off |
| Source systems | SAP and non-SAP databases | Almost any source, including files and web services | SAP Business Suite only | CSV, XLS, XLSX |
| Transformation | Simple, rule based during replication | Rich graphical transformations | Applied by the existing extractor | None |
| Semantics preserved | Table level only | As designed in the job | Yes, business semantics of the extractor | None |
| Setup effort | Medium, needs an SLT server | High, needs a Data Services landscape | Low where extractors exist | Very low |
| Typical use | Operational reporting on live transactions | Consolidating heterogeneous sources | Reusing existing BW extractor logic | Prototyping and small lookup tables |
Each has a dedicated tutorial: SLT, SAP Data Services, DXC, and flat file upload.
Smart Data Access and Smart Data Integration
The built-in list above includes two options that behave unlike the others, because they can avoid copying data at all.
Smart Data Access (SDA). SDA creates a virtual table in SAP HANA that points at a table in a remote system. No data is replicated. When a query touches the virtual table, HANA pushes as much of the work as possible down to the remote database and retrieves only the result. The advantage is that storage cost is zero and the data is always current. The limitation is that query speed now depends on the remote system, so a slow source makes a fast HANA report slow.
Smart Data Integration (SDI). SDI extends the idea with adapters that can also replicate, either in batch or through change data capture, and apply transformations along the way. In effect it brings much of what Data Services offers inside HANA itself, removing the need for a separate ETL landscape on smaller projects.
Choosing between the three families comes down to three questions:
- How current must the data be? Real time points at SLT or SDI change data capture; overnight is enough for Data Services or DXC.
- Does the data need to be copied at all? If it is queried rarely and lives in a capable database, SDA avoids duplication entirely.
- How much transformation is required? Simple filtering suits SLT; complex cleansing, deduplication, and lookups suit Data Services.
Once data has landed, it is exposed to reporting through the information views described in the SAP HANA modeling tutorial.
