Перейти к содержанию

Tantor 7 learned to find problems in PostgreSQL by itself using AI

"Tantor Labs" introduced a new version of the Tantor 7 platform for monitoring and managing PostgreSQL-compatible DBMS. One of the main innovations was AI monitoring: the system independently analyzes the operation of instances, detects potential problems, and suggests options for their elimination and optimization.

A health scale from 1 to 100 has been introduced for assessing the state of databases. The final indicator is calculated based on several parameters and is accompanied by a breakdown of penalty points. This allows administrators to quickly identify the causes of DBMS deterioration without manually comparing a large number of metrics.

Diagnostics can be launched according to a set schedule or automatically during periods of minimal load. The frequency of checks also adapts to the results of previous analyses: problematic instances are checked more often, while stable ones are checked less frequently.

Tantor 7 has expanded configuration management tools. Now, the change history can be filtered and sorted, a node can be selected when rolling back settings, and only individual parameters can be applied. The ability to mass-enable database monitoring has been added, and the limit on simultaneous monitoring of 15 databases has been removed.

In addition, data export to CSV has appeared, and backup, access management, and notifications have been improved. The platform supports DBMS based on PostgreSQL, including Russian commercial forks.