Glossary

Arenadata DB (ADB) is a distributed analytical data storage built on Greengage DB, an open-source massively parallel system.

Arenadata DB Control (ADB Control) is a real-time Arenadata DB query monitoring system.

Arenadata DB Backup Manager (ADBM) is a fault-tolerant system for ADB binary backups management built on pgbackrest.

A type of table where data is optimized for bulk loading. The type is suitable for online analytical processing (OLAP) workloads.

A set of options in Arenadata DB Backup Manager (ADBM) that defines how to manage binary database backups: where to store them, what compression type to use, and so on.

A delivery unit in ADCM designed to facilitate product installation and deployment. It’s an archive containing software packages, configuration file templates, and Ansible tasks that define operations to be performed on the target system.

A system of interconnected nodes comprising a master node and multiple segment nodes, all based on PostgreSQL database instances, that process data in parallel.

A component that facilitates data exchange between ADB and external data sources. For example, ADB ClickHouse Connector, ADB to Kafka Connector, and Kafka to ADB Connector.

In ADBM, a data flow is the process of moving backup data between the source and the target cluster to implement disaster recovery (DR). In ADBM, you can create a data flow configuration to define the backup and restore parameters, such as the restore type (Copy or Streaming) and restore points.

A type of backup, which contains only those database files that have changed since the last full backup was launched.

A mechanism for spreading table rows across multiple segments in a parallel processing architecture. Every table in ADB has a distribution policy.

Arenadata Enterprise Tools or ET is a solution that provides the necessary infrastructure for Arenadata product deployment in environments with limited internet access.

A table definition that points to data stored outside ADB. You can interact with an external table using standard SQL queries as if it were a regular table. External tables use a protocol for connecting to a data source. Internal protocols include FILE, GPFDIST, and HTTP.

Foreign data wrappers (FDW) are libraries that are used by a foreign table to communicate with a remote data source.

A table definition that allows you to access data residing outside the database. Foreign tables use foreign servers to represent remote databases and foreign data wrappers (FDW) to communicate with a remote data source.

A type of backup where the entire database content is backed up.

Grand unified configuration (GUC) is a set of server configuration parameters that allow ADB administrators to manage the DBMS state and configuration.

A type of table where data is stored in an unsorted collection. The type is suitable for online transaction processing (OLTP) workloads.

A type of backup, which contains only those database files that have changed since the last backup of any other type was launched (full, differential, or incremental).

The network that is used for data movement between the master and segments.

The node that acts as the entry point to the database and coordinates segment workloads. It is used by end users to submit queries.

A node that provides high availability support for primary segments. If a primary segment becomes unavailable, the system automatically fails over to the mirror copy.

A set of services (Diamond, Graphite, and Grafana) that monitor a cluster. Clusters that require monitoring should have the Monitoring Clients service installed to be able to integrate with the monitoring cluster.

A technique to split a table into smaller, more manageable parts called partitions in order to increase query performance.

An active node that stores data, receives query plans, and executes queries.

Greenplum Platform Extension Framework (PXF) is a library that allows ADB to query external data sources. PXF includes built-in connectors (such as those allowing you to access HDFS, Hive, and HBase), and users can also create their own connectors. In ADB, PXF is represented as a service that you can manage using the ADCM web interface.

In ADBM, a restore point is a named recovery point, which is the minimum unit of data consistency granularity in ADB clusters.

An instance of the PostgreSQL database that stores and processes its set of data.

A software module that provides specific functionality. Examples of services in an ADB cluster: ADB, ADBM, ADB Control, and connector services.

A component that serves as a warm backup for the primary master instance. It uses write-ahead logging (WAL) to stay synchronized with the primary master.

A logical name for a physical directory on the filesystem where a database can store its data files (such as tables and indexes). It lets you optimize performance by placing frequently accessed data on high-speed storage while storing historical or less critical data on lower-tier storage.

A mechanism that is used in ADBM to distinguish the WAL series generated after the database recovery at the specified restore point from those created before the restore action was applied.

A logical unit that groups one or more SQL statements. These statements are then treated as a single operation — either all of them or none of them succeed. You can monitor transactions on the Monitoring page of ADB Control.

A configuration that links a local database user to a user on an external database to provide access to that database. User mapping is required for using foreign data wrappers (FDW).

Write-ahead logging (WAL) is a technique that ensures data integrity by forcing all database modifications to be written to a persistent, append-only log before the corresponding data pages are updated on disk. WAL plays a central role in point-in-time recovery (PITR) that is used in ADBM.