Yandex Data Processing host classes

The host class determines the computing power allocated for each host in a cluster. When you change the host class for a cluster, all existing hosts change accordingly.

The available storage size does not depend on the host class. For storage limitations, see Quotas and limits.

Available host classesAvailable host classes

Hosts in Yandex Data Processing clusters are deployed on Yandex Compute Cloud VMs. You can create these VMs on any platforms Compute Cloud supports. See Platforms for a detailed description.

A suitable host class depends on driver deploy mode:

• In deployMode=cluster mode, when the driver is deployed on one of the cluster's compute hosts, 4–8 CPU cores and 16 GB RAM are sufficient for the subcluster with the master host.
• In deployMode=client mode, when the driver is deployed on the cluster's master host, the computing power depends on the job logic and the number of concurrently running jobs.

For more information about driver deploy modes and computing resource consumption, see Resource allocation.

The full list of possible host configurations on each platform is provided below.