The SAP HANA Database - An Architecture Overview 论文

摘要

Requirements of enterprise applications have become much more demanding. They require the computation of complex reports on transactional data while thousands of users may read or update records of the same data. The goal of the SAP HANA database is the integration of transactional and analytical workload within the same database management system. To achieve this, a columnar engine exploits modern hardware (multiple CPU cores, large main memory, and caches), compression of database content, maximum parallelization in the database kernel, and database extensions required by enterprise applications, e.g., specialized data structures for hierarchies or support for domain specific languages. In this paper we highlight the architectural concepts employed in the SAP HANA database. We also report on insights gathered with the SAP HANA database in real-world enterprise application scenarios. 1