As the business of a user expands or declines, the peak load of the business varies at different times. Therefore, the database must provide elastic services. OceanBase Database supports elastic services by enabling tenant scaling.
Tenant scaling can improve or reduce the computing and storage capabilities. OceanBase Database provides two methods for tenant scaling: adjusting the resource specifications and adding or removing service nodes.
To improve the service capability of a single node, you can increase the resource specifications (unit config) of the tenant.
For more information, see Scale out or in a tenant by adjusting the resource specifications.
To add service nodes, you can increase the number of units and primary zones. This way, the total service capability of the tenant is improved. In this case, the number of service units increases from N to M, where M > N.
When the number of units of a tenant increases from N to M, each primary zone must have M - N more log streams distributed across M - N new units.
For more information, see Scale out or in a tenant by adjusting the number of units.
When the number of primary zones of a tenant increases from N to M, each unit list must have M - N more log streams, whose leaders are distributed across M - N new primary zones.
For more information, see Scale out or in a tenant by adjusting the number of primary zones.
Tenant scaling in is the reverse of tenant scaling out. It reduces the service capability of a tenant, including the computing and storage capabilities. You can reduce the resource specifications to lower the service capability of a single node, or reduce the number of service nodes by reducing the number of primary zones or units to lower the total service capability of the tenant.