TiKV is intended to fill the role of a unified distributed storage layer. TiKV excels at working with large-scale data by supporting petabyte-scale deployments spanning trillions of rows.

As a graduate project of the , TiKV is originally created by PingCAP to complement .

A TiKV cluster consists of the following components:

TiKV clients let you connect to a TiKV cluster and use raw (simple get/put) API or transaction (with transactional consistency guarantees) API to access and update your data. TiKV clients interact with PD and TiKV through gRPC.

Architecture of TiKV

For more information about the architecture, see .

What’s Next