# Nonclustered Index

![`smart`](NonclusteredIndex03_aa_004_20230922.png)

----------------------------------------------------
The subsequent operations are almost similar to the Clustered Index Table. We need to add the key with the unique identifier of the table (Table ID) to form a unique KEY called 'Table ID + RowID'. This way, a row of data has a global key (KEY) within the entire database scope. The RowID is unique within the entire cluster, and the primary key is unique within the table. These IDs are all of type int64.