<p>The disaggregation of compute and storage in cloud computing environments provide cloud database management and processing systems such as SAP HANA Cloud with new design options like elastic resource scaling that can only be leveraged in the cloud. The elasticity of resources allows for more cost-efficient processing for the system vendor and more cost-control for the consumer. Existing challenges with resource over-provisioning can now be avoided by applying only the resources necessary to the task at hand.</p><p>In this work we describe the architecture and discuss design decisions of HANA’s Elastic Compute Nodes (ECNs) by example of one of SAP’s central business planning applications in the cloud, SAP Integrated Business Planning (SAP IBP), requiring economic viability and scalability of a&#xa0;stateful service, and show SAP IBP’s unique challenges and benefits in cloud environments through ECNs.</p>

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Elastic Compute in SAP HANA Cloud by Example of SAP Integrated Business Planning

  • Boris Gruschko,
  • Kihong Kim,
  • Hyunjun Kim,
  • Taehyung Lee,
  • Michael Mueller,
  • Daniel Ritter

摘要

The disaggregation of compute and storage in cloud computing environments provide cloud database management and processing systems such as SAP HANA Cloud with new design options like elastic resource scaling that can only be leveraged in the cloud. The elasticity of resources allows for more cost-efficient processing for the system vendor and more cost-control for the consumer. Existing challenges with resource over-provisioning can now be avoided by applying only the resources necessary to the task at hand.

In this work we describe the architecture and discuss design decisions of HANA’s Elastic Compute Nodes (ECNs) by example of one of SAP’s central business planning applications in the cloud, SAP Integrated Business Planning (SAP IBP), requiring economic viability and scalability of a stateful service, and show SAP IBP’s unique challenges and benefits in cloud environments through ECNs.