Reserved read/write throughput is an attribute of data tables for high-performance instances. The system reserves resources based on the reserved read/write throughput settings for data tables to meet the throughput demands of the data tables.

When you call the CreateTable operation to create a data table for a high-performance instance, you must specify the reserved read/write throughput. After the data table is created, you can call the UpdateTable operation to modify the reserved read/write throughput settings of the data table.

The reserved read/write throughput is measured by capacity units (CUs). You can set the number of reserved read CUs or write CUs for a data table to a value greater than or equal to 0. By default, the maximum value is 100,000. If you need more than 100,000 reserved read CUs or write CUs, you can submit a ticket to increase the value.

If the number of reserved read CUs or write CUs is greater than 0, you are charged even if no read or write request is sent to the data table.

Read CUs and write CUs are consumed when you call Tablestore API operations to read and write data.

Tablestore generates bills on an hourly basis based on the sum of reserved read CUs and write CUs for all data tables in the high-performance instance. The specified reserved read/write throughput settings may be modified. Tablestore collects statistics about the reserved read CUs and write CUs for data tables at a specific interval, calculates the average number of reserved read CUs and write CUs per hour, and generates bills by multiplying the average number of reserved read CUs and write CUs per hour by their respective unit prices. The unit price of the reserved read CUs or write CUs may change. For more information, see the Pricing tab of Tablestore.