Implementing SAP HANA

December 22, 2011 by Christiane Stagge

There are many deployment scenarios for SAP HANA: It can be used as an accelerator or as a database or as a basis for applications. The SAP HANA appliance comprises software and hardware. In terms of hardware, SAP works with server providers such as HP, IBM, Fujitsu, Cisco, and Dell.

With the help of in-memory technology, SAP HANA can evaluate an extraordinarily high number of data records. To do this, the data or databases are stored in the working memory, from where they can be called immediately, if required. You’ll need a multi-core architecture (8 x 8 CPU), multi-blade servers, data stream speeds of 100 GB/s, and column-based data storage.

SAP HANA is available as a preinstalled hardware appliance. It is provided by SAP and selected SAP partners. These partners are responsible for administration and monitoring, updates, patches, backups, and recovery.

How can you implement SAP HANA in your IT environment?

The system comprises the SAP HANA database, the replication server, database clients, the SAP HANA appliance, the SAP HANA studio, SAP NetWeaver, and the software on which the in-memory technology is based, such as SAP ERP and/or SAP BusinessObjects BI 4.0.

The current release – SAP HANA 1.0 SPS 02 – comprises:

  • The SAP HANA database
  • Database clients
  • The SAP HANA studio (P2 repository)
  • Sybase replication server
  • The SAP HANA load controller
  • SAP host agent
  • LM structure files
  • SAPCAR
  • OS configuration

The SL toolset 1.0 SPS 02 for system administrators comprises:

  • The software update manager for SAP HANA
  • SAP JVM
  • Update manager for SAP HANA

System administrators mainly deal with SAP HANA studio, database clients, Data Service Designer, Information Designer, and Microsoft Excel 2010, while end users work with Microsoft Excel 2010, SAP HANA database clients, and the business intelligence clients.

The data can be replicated using three different technologies:

  • Trigger-based replication
  • ETL-based replication
  • Log-based replication

The following options are available for data loading:

  • Load Controller and Sybase replication server (log-based replication)
  • SAP LT replication (trigger-based replication)
  • Data Services 4.0 (ETL-based replication)

Databases

Among the most common databases on which SAP software is based are MSFT SQL Server Enterprise Edition, Oracle Enterprise Edition, Sybase ASE, IBM DB 2 LUW/UDB, IBM DB/2 zSeries, IBM DB 2 iSeries (previously AS/400), IBM Informix, and SAP MaxDB.

Log file replication is already possible on IBM DB 2 LUW/UDB. It is planned for MSFT SQL Server Enterprise Edition, Oracle Enterprise Edition, and Sybase ASE. Note that it is not possible on IBM DB/2 zSeries, IBM DB 2iSeries, IBM Informix, and SAP MaxDB.

For its data centers, SAP guarantees the highest security and availability. Neither earthquakes nor power failure should pose a danger to the data or the software. All SAP HANA systems are redundant. Within the SAP HANA landscape, there is always a system running on standby.

Hardware for SAP HANA

Currently, the following servers support the SAP HANA appliance:

  • Dell R910, 256 GB RAM
  • Fujitsu RX 600 S5, 128 GB RAM, 256 GB RAM, 512 GB RAM
  • Fujitsu RX 900 S1, 512 GB RAM, 1 TB RAM
  • HP DL 580 G7, 128 GB RAM, 256 GB RAM, 512 GB RAM
  • HP DL 980 G7, 512 GB RAM, 1 TB
  • IBM x3690 X5, 128 GB RAM, 256 GB RAM
  • IBM x3850 X5, 512 GB RAM, 1 TB RAM
  • Cisco UCS C460 M2, 512 GB RAM
  • IBM x3690 X5, 128 GB RAM, 256 GB RAM, 256 GB RAM
  • IBM x3950 X5, 256 GB RAM; 512 GB RAM, 1 TB RAM

Scenarios for SAP HANA

SAP HANA 1.0 supports SAP ERP and SAP NetWeaver Business Warehouse Accelerator. In the next step, SAP HANA for SAP ERP will serve as the basis for SAP NetWeaver Business Warehouse. After that, more applications are planned for SAP HANA. In the more distant future, there are plans for software such as SAP ERP to be fully based on SAP HANA.

If you wish to run BI 4.0 on SAP HANA 1.0 SPS02, you’ll need special licenses for BI 4.0 as well as for the hardware. SAP HANA 1.0 is installed by the hardware providers. The SAP HANA JDBC driver must be manually copied to the BI 4.0 system and updated.

BI 4.0 is still installed on separate hardware. It is already possible for the BI 4.0 clients to access many SAP HANA instances for reporting. There are plans for release 4.1 to support SAP HANA databases. And, in the future, there will be business intelligence appliances for SAP HANA.

SAP HANA in different sizes

The SAP HANA size models are a bit like T-shirt sizes: They range from small to large. Here are the most important facts:

XS

  • 2 x 8 Core Intel Nehalem EX
  • 128 GB working memory
  • 160 GB PCIe flash
  • 1 TB SAS/SSD for data
  • Data decompression from 256 GB to 500 GB
  • Data replication 5 GB/h

S

  • 2 x 8 Core Intel Nehalem EX
  • 256 GB working memory
  • 320 GB PCIe flash
  • 1 TB SAS/SSD
  • Data decompression from 500 GB to 1.25 TB
  • Data replication 5 GB/h

S+

  • 2 x 8 Core Intel Nehalem EX
  • 256 GB working memory
  • 320 GB PCIe flash
  • 1 TB SAS/SSD
  • Data decompression from 500 GB to 2.5 TB
  • Data replication 5 GB/h

M

  • 4 x 8 Core Intel Nehalem EX
  • 512 GB working memory
  • 640 GB PCIe flash
  • 2 TB SAS/SSD
  • Data decompression from 1.25 TB to 2.5 TB
  • Data replication from 5 GB/h to 20 GB/h

M+

  • 4 x 8 Core Intel Nehalem EX
  • 512 GB working memory
  • 640 GB PCIe flash
  • 2 TB SAS/SSD
  • Data decompression from 1.25 TB to 5 TB
  • Data replication from 5 GB/h to 20 GB/h

L

  • 8 x 8 Core Intel Nehalem EX
  • 1 TB working memory
  • 1.2 TB PCIe flash/SSD
  • 4 TB SAS/SSD
  • Data decompression from 2.5 TB to 5 TB
  • Data replication from 5 GB/h to 20 GB/h

Tags: ,

Leave a Reply