GRC
HR
SCM
CRM
BI


Article

 

Maximize Your SAP Deployments with Large Memory Capacity and a Common Server Infrastructure

by Tag Robertson | SAPinsider, Volume 16, Issue 2

April 1, 2015

Many companies are ready to take the leap toward SAP’s latest reporting and transactional capabilities with SAP Business Warehouse and SAP Business Suite 4 SAP HANA (S/4HANA). However, it is critical to ensure your server infrastructure is scalable to handle the growth in data requirements and IT management processes. Read this article to discover how you can best manage the move to SAP S/4HANA with large memory capacity and a common server infrastructure.

 

The SAP HANA database easily integrates online transaction and analytics processing, offering the potential for many new business use cases. Whether it’s the new reporting capabilities with SAP Business Warehouse (SAP BW) and SAP Business Planning and Consolidation, or new transactional capabilities with SAP Business Suite 4 SAP HANA (SAP S/4HANA), companies need to ensure that their server infrastructure is dependable so that they can achieve these capabilities. The server infrastructure must integrate into the existing SAP application landscape, needs to be scalable as data requirements grow, and must incorporate critical IT management processes such as high availability, disaster recovery, and data backup.

While Lenovo may be a new brand in the enterprise server marketplace, the acquisition of the System x server division from IBM — which includes server and services solutions for SAP HANA — brings with it a legacy of enterprise innovation. System x has a strong heritage of SAP HANA deployments and has shipped over 3,600 servers to SAP HANA customers.

Server infrastructure must integrate into the existing SAP application landscape, needs to be scalable as data requirements grow, and must incorporate critical IT management processes such as high availability, disaster recovery, and data backup.

Scaling SAP BW and SAP Business Planning and Consolidation

Depending on the type of SAP application you are planning to deploy, you may have different scaling and availability requirements. The table structures in the application dictate the type of scaling required. SAP BW and SAP Business Planning and Consolidation are structured with row and columnar formatted data. The row data is stored in a master server node and determines the minimum memory sizing requirements for all server nodes in the cluster. The columnar data can be split across multiple server nodes in a cluster configuration. Single as well as multi-node cluster implementations are possible depending on your sizing requirements.

The minimum SAP HANA scale-out cluster configuration consists of three server nodes: one master server node for row store and two worker server nodes for columnar data. This configuration is not fault tolerant. Most IT organizations avoid unplanned outages by implementing an additional standby node for fully automated failover and high availability.

The Lenovo System x certified configurations for SAP HANA use the IBM General Parallel File System (GPFS) and offer the significant scaling and high availability capabilities needed for new business-critical applications, including:

  • Scaling up to two terabytes per node
  • Scaling out to 112 terabytes
  • Integrated high availability and failover
  • Choice of synchronous disaster recovery using GPFS or SAP HANA system replication

When evaluating server infrastructure options for SAP BW or SAP Business Planning and Consolidation, you need to consider several key factors, such as memory scalability, high availability, and failover functionality. Because the sizing of the master node can determine the sizing for each server node in the cluster, it is important to be able to scale up the memory of the master node.

SAP determines the server sizing and memory requirements of its applications based on a number of factors, including memory per processor core ratios. Based on the current ratios, the maximum memory capacity for each node in the eight-socket Lenovo System x3950 X6 server based on Intel Xeon Processor E7v2 is two terabytes.

With GPFS, you can scale out to 112 terabytes of SAP HANA memory capacity. In addition, you can set up high availability nodes that handle auto host failover if a server node in the cluster goes down. Large cluster configurations can implement failure groups that allow the cluster to be segmented into groups of nodes. A standby node can then be assigned to a specific segment in the cluster.   

There are still some SAP BW customers that may not be ready to make the move to SAP HANA because they cannot migrate to SAP BW 7.3, a prerequisite for moving to SAP HANA. For these customers, the only option for real-time analytics is SAP BW Accelerator. Lenovo offers the latest server configurations for SAP BW Accelerator, enabling customers to access real-time analytics in their current environment. The latest configurations are based on the Lenovo Flex System x240 compute node and IBM System Storage v3700. These configurations are delivered as an integrated solution and can quickly be inserted into existing SAP BW environments. There is also an option for a single server configuration based on Lenovo System x3850 X6.

Scaling SAP S/4HANA

The scaling requirements for SAP S/4HANA are different from those for SAP BW and SAP Business Planning and Consolidation because SAP S/4HANA has very large tables and requires a large, contiguous memory space. As a result, these applications require a large symmetric multi-processing capability with large memory capacity. As the database grows, it is critical to have the ability to scale up the server infrastructure to keep pace with data growth.

Lenovo System x and Flex System servers have the ability to support eight processors with 15 cores in each Intel Xeon E7v2 processor, providing up to six terabytes of memory for SAP HANA applications in production environments and up to eight processors and 12 terabytes of memory for non-production environments.

Choosing server infrastructure that can seamlessly scale from single-server configurations to multi-server cluster configurations provides investment protection by allowing you to deploy the same server platform in development, test, and production environments for SAP BW or SAP S/4HANA.

Operational Considerations

In addition to having a server with large memory capacity, it is important to have common server infrastructure for development, test, quality assurance, and production environments. To simplify management and reduce operational costs, organizations should standardize around a platform that can both scale up and scale out memory, so that common system management practices can be established to maintain and update business-critical systems for SAP HANA.

Choosing server infrastructure that can seamlessly scale from single-server configurations to multi-server cluster configurations provides investment protection by allowing you to deploy the same server platform in development, test, and production environments. This further enables a common software stack and server management policy.

The Lenovo System x configurations with GPFS can support multiple generations of SAP HANA server technology in a single cluster. In addition, if your IT organization is resource constrained, Lenovo Enterprise Solution Services can help you manage and maintain your SAP HANA server infrastructure. 

The Lenovo System x solution for SAP HANA can also help customers that have strict data security and management requirements. The native cryptographic capabilities of the advanced version of GPFS include the secure deletion of data. This ensures data confidentiality in a multisite disaster recovery implementation because all at-rest and in-flight data between data centers is encrypted, and because only encrypted SAP HANA data leaves the data center.

This is ideal for installations using shared leased lines or public internet connections for their high availability and disaster recovery setup. It also helps prevent data leakage when SAP HANA servers are improperly decommissioned.

Learn More

Applications such as SAP BW powered by SAP HANA and SAP Business Planning and Consolidation require a scale-out architecture to support large memory requirements, while SAP S/4HANA typically requires a large scale-up memory architecture. Choosing server infrastructure that can seamlessly scale from single-server configurations to multi-server cluster configurations provides investment protection by allowing you to deploy the same server platform in development, test, and production environments for SAP BW or SAP S/4HANA, enabling you to have a common server platform, software stack, and server management policy across all of your SAP application environments.

Lenovo’s System x enterprise heritage and experience in installing and managing SAP HANA systems can help you speed your SAP HANA deployment and realize the business potential that new SAP HANA applications enable. For more information, visit www.lenovo.com/servers.

An email has been sent to:





 

Tag Robertson
Tag Robertson

Senior Product and Alliance Manager
Lenovo



More from SAPinsider



COMMENTS

Please log in to post a comment.

SAPinsider
FAQ