Basically question is in the topic. Isn’t explained.
Course: An Introduction to Apache Airflow - Learn Interactively
Lesson: https://www.educative.io/courses/introduction-to-apache-airflow/hooks-and-connections
Hi @Dmitry_Polovinkin !
A metastore database is used to store connections. The metastore database, which is a part of the Airflow metadata database, serves as a centralized repository for various Airflow-related metadata including connections, DAGs, tasks, variables, and more. The metastore database is typically a relational database management system (RDBMS) like SQLite, MySQL, or PostgreSQL. It allows Airflow to persistently store and manage information related to workflows, facilitating tasks like logging, tracking, and managing workflow execution.
When it comes to connections, storing them in the metastore database means that the information required to establish connections with external systems, such as databases, cloud services, or other systems, is securely stored within the Airflow metadata database. This approach provides a centralized and secure way to manage sensitive connection information, including credentials, hostnames, and other connection-specific details, without hardcoding them directly into the DAG code.
By storing connections in the metastore database, Airflow provides a unified interface to access and manage these connections. This enables users to easily reference and utilize connections across different tasks within a DAG without the need for redundant connection configurations. Additionally, this centralized approach simplifies the management of connections and enhances the security of sensitive information, ensuring that the necessary credentials are not exposed within the DAG code, but are instead securely stored and managed within the Airflow metadata database.
I hope it helps. Happy Learning 
1 Like