IB Computer Science Topic A3.4: Alternative Databases and Data Warehouses Key Definitions

The tools and practices an organisation uses to query its stored data and turn the results into decisions, covering decision support, trend analysis, forecasting and performance monitoring.

A database hosted and run on a provider's remote servers rather than on hardware the organisation owns. The provider supplies capacity on demand and handles backups, patching, security and maintenance.

Clustering is a technique used to group a set of objects so that objects in the same group (or cluster) are more similar to each other than to those in other groups.

The storing of copies of the same data at more than one site in a distributed database, which improves read performance and availability at the cost of keeping every copy up to date.

A central store that collects historical data from an organisation's operational systems and holds it in a form designed for analysis rather than for running the business.

Deadlock occurs when processes are stuck waiting for resources held by each other.

A database whose data is stored across several physical sites, often on separate machines or in separate locations, while presenting itself to users and applications as a single logical database.

The property of a distributed database that hides how and where data is stored, so users and applications work with it as one database without knowing which site holds any given item.

The scheduled three-stage process that fills a data warehouse: data is extracted from each source system, transformed into the warehouse's agreed formats, and loaded in as new rows.

A database that holds its working data in RAM rather than on disk, so reads and writes avoid the slowest step in the storage hierarchy. The contents are lost when power is lost unless a disk log or non-volatile memory is used.

A database that stores data without a fixed relational schema, so records in the same collection can carry different fields. The main families are document stores, key-value stores, column-family stores and graph databases.

Online analytical processing (OLAP) is the analysis of historical data already held in a data warehouse, organised into pre-aggregated multidimensional views so the same measure can be summarised along dimensions such as time, region and product.

Online transaction processing (OLTP) is the processing of the many small individual transactions that run an organisation day to day, such as recording a single sale, where each write must complete in milliseconds.

The ability of a system to handle more data and more users by adding further nodes rather than by replacing existing hardware with something more powerful.

A database that stores geometry such as points, lines and polygons as a data type and indexes rows by position. It answers queries about distance, overlap and containment directly.

A protocol that makes a transaction atomic across several sites. A coordinator first asks every participating site to vote on whether it can commit, then tells every site to commit only if all votes were yes, and to abort otherwise.