OLAP with a Database Cluster

This chapter presents a new approach to online decision support systems that is scalable, fast, and capable of analysing up-to-date data. It is based on a database cluster: a cluster of commercial off-the-shelf computers as hardware infrastructure and off-the-shelf database management systems as transactional storage managers. We focus on central architectural issues and on the performance implications of such a cluster-based decision support system. In the first half, we present a scalable infrastructure and discuss physical data design alternatives for cluster-based online decision support systems. In the second half of the chapter, we discuss query routing algorithms and freshness-aware scheduling. This protocol enables users to seamlessly decide how fresh the data analysed should be by allowing for different degrees of freshness of the online analytical processing (OLAP) nodes. In particular it becomes then possible to trade freshness of data for query performance. INtrODUctION Online analytical processing (OLAP) systems must cope with huge volumes of data and at the same time must allow for short response times to facilitate interactive usage. They must also be capable to scale, meaning to be easily extensible with the increasing data volumes accumulated. Furthermore, the requirement that the data analysed should be up-to-date is becoming more and more important. However, not only are these contrary requirements, but they also run counter to the performance needs of the day-to-day business. Most OLAP systems nowadays are kept separated from mission critical systems. This means that they offer a compromise between “up-todateness,” that is, freshness (or currency) of data, and query response times. The data needed are propagated into the OLAP system on a regular basis, preferably when not slowing down dayto-day business, for example, during nights or weekends. OLAP users have no alternative but to analyse stale data.