Jump toUpdate content

Serverless SQL Databases - Concepts

Active state

A database is in an active state when it receives queries frequently (at least once every 5 minutes). If it receives no query for 5 minutes, the database switches to an idle state, and the billing for compute capacity stops.


Autoscaling can automatically scale your database resources up or down, depending on vCPU consumption.


A database backup is a copy of the data stored in the database (including data structure, such as tables). Database backups in Serverless SQL Database are logical backups (as opposed to raw disk volume backups, also called “snapshots”). Database backups can be restored in the current database or any other PostgreSQL database and allows to recover from database crash, data corruption, or data loss.

Cold start

Cold start is the time an idle database takes to handle the first request. It usually takes less than five seconds.

Connection pooling

Connection pooling is the process of mutualizing client connections used to connect to the database. By default, connections to PostgreSQL databases use a lot of memory, which can have a significant impact on performance when hundreds of clients are connected to a database at the same time. With connection pooling, a common component (a “connection pooler”, such as PgBouncer) connects to the database and maintains multiple connections to it. Clients then connect to the connection pooler instead of connecting directly to the database, allowing for better performance and more concurrent connections.


Credentials are the login and password (also called “username” and “password” in PostgreSQL vocabulary) used by a user or an application to connect to a database. With Scaleway Serverless SQL Databases, logins are IAM principal IDs and passwords are IAM secret keys.

Database engine

A database engine is the software that stores and retrieves your data from a database. Serverless SQL databases use the PostgreSQL engine.

High availability

Databases can become unavailable due to several issues (host crashes, disk corruption, network unavailability, etc.). High Availability ensures that databases stay available most of the time and recover quickly from issues so that applications relying on them remain available as well.


A database switches to an idle state if it receives no query for 5 minutes. You are not billed for compute resources while your database is in an idle state.


Logs register the activity of your database. They provide useful information for debugging or to know more about the behavior and activity of your databases.


PostgreSQL is an open source object-relational database system known for its reliability, robustness, and performance.

Region and AZ

A Region is as a geographical area such as France (Paris: fr-par) or the Netherlands (Amsterdam: nl-ams) in which Scaleway products and resources are located. It can contain multiple Availability Zones.

An Availability Zone refers to the geographical location within a region, such as waw-1(Warsaw, Poland), in which your Scaleway resource will be created. The latency between multiple AZs of the same region is low, as they have a common network layer.

For an extensive list of which regions and AZ a resource is available in, refer to our Products availability guide

Relational database

A database type that relies on the relational model and that can be queried using the SQL language. In the relational model, different data objects are stored in different tables, which can be linked by common IDs, or keys. Relational databases provide an intuitive, straightforward way of representing data in tables, which favors data normalization, and provide the strongest data consistency guarantees (compared to NoSQL alternatives).

Serverless database

A serverless database is a fully managed database-as-a-service (DBaaS) that allocates and scales compute and storage resources automatically to match the incoming workload.


The status of a database changes according to its activity and provides information on ongoing operations.