# Clustered Vs. Non-Clustered Tables (Continue...)

+ non-clustered tables
  + A table whose row data storage order does not necessarily match the order in which the primary key is stored is a nonclustered table
  + The key of the row data consists of _tidb_rowid implicitly allocated within TiDB, and the primary key is essentially a unique index
  + When you access a row records through the primary key, you cannot get the row records directly. You need to get the _tidb_rowid of the row from the additional stored primary key first, and then get the row record through the _tidb_rowid of this row. This will return the table one more time than a clustered table.
```
CREATE TABLE t (a BIGINT PRIMARY KEY NONCLUSTERED, b VARCHAR(255));
```
------------------------------------------------------------
- Non-clustered tables
  - A table whose row data storage order does not necessarily match the order in which the primary key is stored is a non-clustered table
  - The key of the row data consists of _tidb_rowid implicitly allocated within TiDB, and the primary key is essentially a unique index
  - When you access a row records through the primary key, you cannot get the row records directly. You need to get the _tidb_rowid of the row from the extra stored primary key first, and then get the row record through the row's _tidb_rowid. This will return the table one more time than a clustered table
CREATE TABLE t (a BIGINT PRIMARY KEY NONCLUSTERED, b VARCHAR (255));