期刊名称：Bulletin of the Technical Committee on Data Engineering
出版年度：2013
卷号：36
期号：2
出版社：IEEE Computer Society
摘要：Two emerging hardware trends have re-initiated the development of in-core database systems: everincreasing main-memory capacities and vast multi-core parallel processing power. Main-memory ca-pacities of several TB allow to retain all transactional data of even the largest applications in-memoryon one (or a few) servers. The vast computational power in combination with low data managementoverhead yields unprecedented transaction performance which allows to push transaction processing(away from application servers) into the database server and still "leaves room" for additional queryprocessing directly on the transactional data. Thereby, the often postulated goal of real-time businessintelligence, where decision makers have access to the latest version of the transactional state, becomesfeasible. In this paper we will survey the HyPerScript transaction programming language, the main-memory indexing technique ART, which is decisive for high transaction processing performance, andHyPer's transaction management that allows heterogeneous workloads consisting of short pre-cannedtransactions, OLAP-style queries, and long interactive transactions
