Terms and definitions

Database

A database (DB) is a dedicated isolated part of the YDB service resources. The YDB service provides database scalability and fault-tolerance, as well as automatic data replication.

Regions and availability zones

The YDB database is regional and is located in the ru-central region. The ru-central region has three availability zones. For more information about regions and availability zones, see Availability zones.

Directories

To easily organize databases, YDB lets you create directories like in the file system.

Computing resources

Computing resources are combinations of vCPUs and RAM. In YDB, they're provided based on Yandex Compute Cloud VMs. You can create these VMs on any of the platforms that Compute Cloud supports. For a detailed description of the platforms, see Platforms.

Types of computing resources:

Platform	Computing resource	Number of vCPUs	Guaranteed vCPU performance	RAM, GB
Intel Cascade Lake	s2.medium	8	100%	32

To increase database throughput, you can increase the amount of computing resources.

Storage groups

A storage group is a redundant array of independent disks that are networked in a single logical unit. Storage groups increase fault tolerance through redundancy and improve performance.

In YDB, each of the three availability zones uses three disks. This configuration ensures stability even if a zone or disk fails. It also provides a redundancy factor of 3.

A storage group may consist of any of the following types of storage devices:

• SSD (network-ssd): A fast network drive. Network block storage on an SSD. Suitable for high-performance OLTP applications.
• HDD (network-hdd): A standard network drive. Network block storage on an HDD. Suitable for storing large volumes (petabytes) of cold data.

YDB lets you allocate additional storage groups as your data grows.