Arenadata DB

Partitioning is a way to increase query performance by logical dividing large tables (such as fact tables) into smaller, more manageable parts called partitions. Due to partitioning query optimizers can read the limited number of table rows (based on predicate conditions) instead of scanning all the table contents each time.

Distribution is one of the most important concepts underlying ADB. It means storing table rows on different segments of the cluster. The more uniform the distribution of the data between segments is — the higher the whole cluster performance is.

ADB supports several types of tables with different data storage models: heap, append-optimized row-oriented, append-optimized column-oriented. The article describes features and possible use cases for each table type.

Table is a main data object in ADB. ADB tables are similar to tables in any relational database, except distribution of table rows across the different cluster segments. ADB provides several table storage types and compression options that can be used to store data in the most efficient way.

Learn about software requirements for the Arenadata DB cluster installation.

Spill files (also known as work files) are created on disks if there is not enough memory to store temporary data when executing data queries. This temporary data may include JOIN hashes, sorting results, redistributed tables, etc.

Arenadata DB Control (ADB Control) is a real-time Arenadata DB query monitoring system. Learn about ADB Control features, architecture, and some internal details.

This article describes the parameters that can be configured for ADB services via ADCM.

Learn about hardware requirements (CPU, RAM, disk) for installation of ADB, ADB Control, ADBM.

Learn how to configure and use backup configurations in ADBM.