A Microsoft-owned Citus Data has released the distributed DBMS Citus 13.0, implemented as an extension to PostgreSQL 17. Citus enables horizontal scaling of PostgreSQL in a cluster based on standard hardware and allows data to be distributed across nodes through sharding, with column-level partitioning and storage schema settings. For applications, the Citus cluster appears as one large PostgreSQL server, combining resources from its constituent nodes. The code is written in C and is distributed under the AGPLv3 license.
Sharding allows for the organization of storage for a very large volume of data, the total size of which significantly exceeds the local storage of each cluster node. Individual tables can be forcibly replicated across all nodes to speed up merge operations and external key operations. To save disk space, data distributed across different nodes can be stored in a compressed format.
Distributed queries can be sent to any node in the cluster, but the management and modification of the data schema must only be performed through the node coordinating the cluster's operation. Incoming requests from clients are distributed across the necessary nodes, and if they span multiple nodes, their processing is parallelized. The cluster can be expanded as the size of the stored data grows by adding additional nodes and initiating rebalancing.
Typical usage examples of Citus include executing analytical queries and processing large data volumes in the form of time series (for example, logs or sensor state surveys). Citus is also suitable for modernizing existing infrastructures based on a single a szerver PostgreSQL, whose performance and storage have become insufficient due to increased load or data volume. With the Patroni toolkit, fault-tolerant configurations can be created with replicated backup nodes capable of taking the place of primary nodes in case of failure.
Changes in the release of Citus 13.0:
- Transition to using PostgreSQL 17 branch.
- A JSON_TABLE() function can be used in distributed queries to convert data from JSON format into a relational representation that can be worked with like regular PostgreSQL tables.
- Support for the syntax 'MERGE … WHEN NOT MATCHED BY SOURCE'.
- The ability to define the access method for distributed partitioned tables using the 'CREATE TABLE … USING' expression, as well as changing the access method via 'ALTER TABLE … SET ACCESS METHOD'.
- Support has been added for specifying automatically generated identification columns in distributed partitioned tables.
- The application of exclusion constraints in distributed partitioned tables has been permitted.
- Issues with role synchronization between nodes have been resolved.
- Data distribution for sharding has been improved when adding new nodes to the cluster.
- The 'EXPLAIN' command has been enhanced with 'MEMORY' and 'SERIALIZE' options, allowing for the determination of memory size used and the time spent on data transformation for network transmission.
Forrás: opennet.ru
