Bewerken

Delen via


CREATE TABLE (Transact-SQL)

Applies to: SQL Server Azure SQL Database Azure SQL Managed Instance SQL database in Microsoft Fabric

Creates a new table in the database.

Note

For reference to Warehouse in Microsoft Fabric, visit CREATE TABLE (Fabric Data Warehouse). For reference to Azure Synapse Analytics and Analytics Platform System (PDW), visit CREATE TABLE (Azure Synapse Analytics).

Transact-SQL syntax conventions

Syntax options

Common syntax

Simple CREATE TABLE syntax (common if not using options):

CREATE TABLE
    { database_name.schema_name.table_name | schema_name.table_name | table_name }
    ( { <column_definition> } [ ,... n ] )
[ ; ]

Full syntax

Disk-based CREATE TABLE syntax:

CREATE TABLE
    { database_name.schema_name.table_name | schema_name.table_name | table_name }
    [ AS FileTable ]
    ( { <column_definition>
        | <computed_column_definition>
        | <column_set_definition>
        | [ <table_constraint> ] [ ,... n ]
        | [ <table_index> ] }
          [ ,... n ]
          [ PERIOD FOR SYSTEM_TIME ( system_start_time_column_name
             , system_end_time_column_name ) ]
      )
    [ ON { partition_scheme_name ( partition_column_name )
           | filegroup
           | "default" } ]
    [ TEXTIMAGE_ON { filegroup | "default" } ]
    [ FILESTREAM_ON { partition_scheme_name
           | filegroup
           | "default" } ]
    [ WITH ( <table_option> [ ,... n ] ) ]
[ ; ]

<column_definition> ::=
column_name <data_type>
    [ FILESTREAM ]
    [ COLLATE collation_name ]
    [ SPARSE ]
    [ MASKED WITH ( FUNCTION = 'mask_function' ) ]
    [ [ CONSTRAINT constraint_name ] DEFAULT constant_expression ]
    [ IDENTITY [ ( seed , increment ) ]
    [ NOT FOR REPLICATION ]
    [ GENERATED ALWAYS AS { ROW | TRANSACTION_ID | SEQUENCE_NUMBER } { START | END } [ HIDDEN ] ]
    [ [ CONSTRAINT constraint_name ] {NULL | NOT NULL} ]
    [ ROWGUIDCOL ]
    [ ENCRYPTED WITH
        ( COLUMN_ENCRYPTION_KEY = key_name ,
          ENCRYPTION_TYPE = { DETERMINISTIC | RANDOMIZED } ,
          ALGORITHM = 'AEAD_AES_256_CBC_HMAC_SHA_256'
        ) ]
    [ <column_constraint> [ ,... n ] ]
    [ <column_index> ]

<data_type> ::=
[ type_schema_name. ] type_name
    [ ( precision [ , scale ] | max |
        [ { CONTENT | DOCUMENT } ] xml_schema_collection ) ]

<column_constraint> ::=
[ CONSTRAINT constraint_name ]
{
   { PRIMARY KEY | UNIQUE }
        [ CLUSTERED | NONCLUSTERED ]
        [ ( <column_name> [ ,... n ] ) ]
        [
            WITH FILLFACTOR = fillfactor
          | WITH ( <index_option> [ ,... n ] )
        ]
        [ ON { partition_scheme_name ( partition_column_name )
            | filegroup | "default" } ]

  | [ FOREIGN KEY ]
        REFERENCES [ schema_name. ] referenced_table_name [ ( ref_column ) ]
        [ ON DELETE { NO ACTION | CASCADE | SET NULL | SET DEFAULT } ]
        [ ON UPDATE { NO ACTION | CASCADE | SET NULL | SET DEFAULT } ]
        [ NOT FOR REPLICATION ]

  | CHECK [ NOT FOR REPLICATION ] ( logical_expression )
}

<column_index> ::=
 INDEX index_name [ CLUSTERED | NONCLUSTERED ]
    [ WITH ( <index_option> [ ,... n ] ) ]
    [ ON { partition_scheme_name ( column_name )
         | filegroup_name
         | default
         }
    ]
    [ FILESTREAM_ON { filestream_filegroup_name | partition_scheme_name | "NULL" } ]

<computed_column_definition> ::=
column_name AS computed_column_expression
[ PERSISTED [ NOT NULL ] ]
[
    [ CONSTRAINT constraint_name ]
    { PRIMARY KEY | UNIQUE }
        [ CLUSTERED | NONCLUSTERED ]
        [
            WITH FILLFACTOR = fillfactor
          | WITH ( <index_option> [ ,... n ] )
        ]
        [ ON { partition_scheme_name ( partition_column_name )
        | filegroup | "default" } ]

    | [ FOREIGN KEY ]
        REFERENCES referenced_table_name [ ( ref_column ) ]
        [ ON DELETE { NO ACTION | CASCADE } ]
        [ ON UPDATE { NO ACTION } ]
        [ NOT FOR REPLICATION ]

    | CHECK [ NOT FOR REPLICATION ] ( logical_expression )
]

<column_set_definition> ::=
column_set_name XML COLUMN_SET FOR ALL_SPARSE_COLUMNS

<table_constraint> ::=
[ CONSTRAINT constraint_name ]
{
    { PRIMARY KEY | UNIQUE }
        [ CLUSTERED | NONCLUSTERED ]
        ( column_name [ ASC | DESC ] [ ,... n ] )
        [
            WITH FILLFACTOR = fillfactor
           | WITH ( <index_option> [ ,... n ] )
        ]
        [ ON { partition_scheme_name (partition_column_name)
            | filegroup | "default" } ]
    | FOREIGN KEY
        ( column_name [ ,... n ] )
        REFERENCES referenced_table_name [ ( ref_column [ ,... n ] ) ]
        [ ON DELETE { NO ACTION | CASCADE | SET NULL | SET DEFAULT } ]
        [ ON UPDATE { NO ACTION | CASCADE | SET NULL | SET DEFAULT } ]
        [ NOT FOR REPLICATION ]
    | CHECK [ NOT FOR REPLICATION ] ( logical_expression )
}

<table_index> ::=
{
    {
      INDEX index_name [ UNIQUE ] [ CLUSTERED | NONCLUSTERED ]
         ( column_name [ ASC | DESC ] [ ,... n ] )
    | INDEX index_name CLUSTERED COLUMNSTORE [ ORDER (column_name [ , ...n ] ) ]
    | INDEX index_name [ NONCLUSTERED ] COLUMNSTORE ( column_name [ ,... n ] )
    }
    [ INCLUDE ( column_name [ ,... n ] ) ]
    [ WHERE <filter_predicate> ]
    [ WITH ( <index_option> [ ,... n ] ) ]
    [ ON { partition_scheme_name ( column_name )
         | filegroup_name
         | default
         }
    ]
    [ FILESTREAM_ON { filestream_filegroup_name | partition_scheme_name | "NULL" } ]
}

<table_option> ::=
{
    [ DATA_COMPRESSION = { NONE | ROW | PAGE }
      [ ON PARTITIONS ( { <partition_number_expression> | <range> }
      [ ,... n ] ) ] ]
    [ XML_COMPRESSION = { ON | OFF }
      [ ON PARTITIONS ( { <partition_number_expression> | <range> }
      [ ,... n ] ) ] ]
    [ FILETABLE_DIRECTORY = <directory_name> ]
    [ FILETABLE_COLLATE_FILENAME = { <collation_name> | database_default } ]
    [ FILETABLE_PRIMARY_KEY_CONSTRAINT_NAME = <constraint_name> ]
    [ FILETABLE_STREAMID_UNIQUE_CONSTRAINT_NAME = <constraint_name> ]
    [ FILETABLE_FULLPATH_UNIQUE_CONSTRAINT_NAME = <constraint_name> ]
    [ SYSTEM_VERSIONING = ON
        [ ( HISTORY_TABLE = schema_name.history_table_name
          [ , DATA_CONSISTENCY_CHECK = { ON | OFF } ]
    ) ]
    ]
    [ REMOTE_DATA_ARCHIVE =
      {
        ON [ ( <table_stretch_options> [ ,... n] ) ]
        | OFF ( MIGRATION_STATE = PAUSED )
      }
    ]
    [ DATA_DELETION = ON
          { (
             FILTER_COLUMN = column_name,
             RETENTION_PERIOD = { INFINITE | number { DAY | DAYS | WEEK | WEEKS
                              | MONTH | MONTHS | YEAR | YEARS }
        ) }
    ]
    [ LEDGER = ON [ ( <ledger_option> [ ,... n ] ) ]
    | OFF
    ]
}

<ledger_option>::=
{
    [ LEDGER_VIEW = schema_name.ledger_view_name [ ( <ledger_view_option> [ ,... n ] ) ] ]
    [ APPEND_ONLY = ON | OFF ]
}

<ledger_view_option>::=
{
    [ TRANSACTION_ID_COLUMN_NAME = transaction_id_column_name ]
    [ SEQUENCE_NUMBER_COLUMN_NAME = sequence_number_column_name ]
    [ OPERATION_TYPE_COLUMN_NAME = operation_type_id column_name ]
    [ OPERATION_TYPE_DESC_COLUMN_NAME = operation_type_desc_column_name ]
}

<table_stretch_options> ::=
{
    [ FILTER_PREDICATE = { NULL | table_predicate_function } , ]
      MIGRATION_STATE = { OUTBOUND | INBOUND | PAUSED }
 }

<index_option> ::=
{
    PAD_INDEX = { ON | OFF }
  | FILLFACTOR = fillfactor
  | IGNORE_DUP_KEY = { ON | OFF }
  | STATISTICS_NORECOMPUTE = { ON | OFF }
  | STATISTICS_INCREMENTAL = { ON | OFF }
  | ALLOW_ROW_LOCKS = { ON | OFF }
  | ALLOW_PAGE_LOCKS = { ON | OFF }
  | OPTIMIZE_FOR_SEQUENTIAL_KEY = { ON | OFF }
  | COMPRESSION_DELAY = { 0 | delay [ Minutes ] }
  | DATA_COMPRESSION = { NONE | ROW | PAGE | COLUMNSTORE | COLUMNSTORE_ARCHIVE }
       [ ON PARTITIONS ( { partition_number_expression | <range> }
       [ ,... n ] ) ]
  | XML_COMPRESSION = { ON | OFF }
      [ ON PARTITIONS ( { <partition_number_expression> | <range> }
      [ ,... n ] ) ]
}

<range> ::=
<partition_number_expression> TO <partition_number_expression>

Syntax for memory optimized tables

Memory optimized CREATE TABLE syntax:

CREATE TABLE
    { database_name.schema_name.table_name | schema_name.table_name | table_name }
    ( { <column_definition>
    | [ <table_constraint> ] [ ,... n ]
    | [ <table_index> ]
      [ ,... n ] }
      [ PERIOD FOR SYSTEM_TIME ( system_start_time_column_name
        , system_end_time_column_name ) ]
)
    [ WITH ( <table_option> [ ,... n ] ) ]
 [ ; ]

<column_definition> ::=
column_name <data_type>
    [ COLLATE collation_name ]
    [ GENERATED ALWAYS AS ROW { START | END } [ HIDDEN ] ]
    [ NULL | NOT NULL ]
    [ [ CONSTRAINT constraint_name ] DEFAULT memory_optimized_constant_expression ]
    | [ IDENTITY [ ( 1, 1 ) ] ]
    [ <column_constraint> ]
    [ <column_index> ]

<data_type> ::=
 [type_schema_name. ] type_name [ (precision [ , scale ]) ]

<column_constraint> ::=
 [ CONSTRAINT constraint_name ]
{
  { PRIMARY KEY | UNIQUE }
      { NONCLUSTERED
        | NONCLUSTERED HASH WITH ( BUCKET_COUNT = bucket_count )
      }
  [ ( <column_name> [ ,... n ] ) ]
  | [ FOREIGN KEY ]
        REFERENCES [ schema_name. ] referenced_table_name [ ( ref_column ) ]
  | CHECK ( logical_expression )
}

<table_constraint> ::=
 [ CONSTRAINT constraint_name ]
{
   { PRIMARY KEY | UNIQUE }
     {
       NONCLUSTERED ( column_name [ ASC | DESC ] [ ,... n ])
       | NONCLUSTERED HASH ( column_name [ ,... n ] ) WITH ( BUCKET_COUNT = bucket_count )
                    }
    | FOREIGN KEY
        ( column_name [ ,... n ] )
        REFERENCES referenced_table_name [ ( ref_column [ ,... n ] ) ]
    | CHECK ( logical_expression )
}

<column_index> ::=
  INDEX index_name
{ [ NONCLUSTERED ] | [ NONCLUSTERED ] HASH WITH ( BUCKET_COUNT = bucket_count ) }

<table_index> ::=
  INDEX index_name
{   [ NONCLUSTERED ] HASH ( column_name [ ,... n ] ) WITH ( BUCKET_COUNT = bucket_count )
  | [ NONCLUSTERED ] ( column_name [ ASC | DESC ] [ ,... n ] )
      [ ON filegroup_name | default ]
  | CLUSTERED COLUMNSTORE [ WITH ( COMPRESSION_DELAY = { 0 | delay [ Minutes ] } ) ]
      [ ON filegroup_name | default ]

}

<table_option> ::=
{
    MEMORY_OPTIMIZED = ON
  | DURABILITY = { SCHEMA_ONLY | SCHEMA_AND_DATA }
  | SYSTEM_VERSIONING = ON [ ( HISTORY_TABLE = schema_name.history_table_name
        [, DATA_CONSISTENCY_CHECK = { ON | OFF } ] ) ]

}

Arguments

database_name

The name of the database in which the table is created. database_name must specify the name of an existing database. If not specified, database_name defaults to the current database. The login for the current connection must be associated with an existing user ID in the database specified by database_name, and that user ID must have CREATE TABLE permissions.

schema_name

The name of the schema to which the new table belongs.

table_name

The name of the new table. Table names must follow the rules for identifiers. table_name can be a maximum of 128 characters, except for local temporary table names (names prefixed with a single number sign (#)) that can't exceed 116 characters.

AS FileTable

Applies to: SQL Server 2012 (11.x) and later.

Creates the new table as a FileTable. You don't specify columns because a FileTable has a fixed schema. For more information, see FileTables.

column_name AS computed_column_expression

An expression that defines the value of a computed column. A computed column is a virtual column that isn't physically stored in the table, unless the column is marked PERSISTED. The column is computed from an expression that uses other columns in the same table. For example, a computed column can have the definition: cost AS price * qty. The expression can be a noncomputed column name, constant, function, variable, and any combination of these connected by one or more operators. The expression can't be a subquery or contain alias data types.

Computed columns can be used in select lists, WHERE clauses, ORDER BY clauses, or any other locations in which regular expressions can be used, with the following exceptions:

  • Computed columns must be marked PERSISTED to participate in a FOREIGN KEY or CHECK constraint.

  • A computed column can be used as a key column in an index or as part of any PRIMARY KEY or UNIQUE constraint, if the computed column value is defined by a deterministic expression and the data type of the result is allowed in index columns.

    For example, if the table has integer columns a and b, the computed column a + b may be indexed, but computed column a + DATEPART(dd, GETDATE()) can't be indexed because the value may change in subsequent invocations.

  • A computed column can't be the target of an INSERT or UPDATE statement.

Note

Each row in a table can have different values for columns that are involved in a computed column; therefore, the computed column may not have the same value for each row.

Based on the expressions that are used, the nullability of computed columns is determined automatically by the Database Engine. The result of most expressions is considered nullable even if only nonnullable columns are present, because possible underflows or overflows also produce NULL results. Use the COLUMNPROPERTY function with the AllowsNull property to investigate the nullability of any computed column in a table. An expression that is nullable can be turned into a nonnullable one by specifying ISNULL with the check_expression constant, where the constant is a nonnull value substituted for any NULL result. REFERENCES permission on the type is required for computed columns based on common language runtime (CLR) user-defined type expressions.

PERSISTED

Specifies that the SQL Server Database Engine will physically store the computed values in the table, and update the values when any other columns on which the computed column depends are updated. Marking a computed column as PERSISTED lets you create an index on a computed column that is deterministic, but not precise. For more information, see Indexes on Computed Columns. Any computed columns that are used as partitioning columns of a partitioned table must be explicitly marked PERSISTED. computed_column_expression must be deterministic when PERSISTED is specified.

ON { partition_scheme | filegroup | "default" }

Specifies the partition scheme or filegroup on which the table is stored. If partition_scheme is specified, the table is to be a partitioned table whose partitions are stored on a set of one or more filegroups specified in partition_scheme. If filegroup is specified, the table is stored in the named filegroup. The filegroup must exist within the database. If "default" is specified, or if ON isn't specified at all, the table is stored on the default filegroup. The storage mechanism of a table as specified in CREATE TABLE can't be subsequently altered.

ON { partition_scheme | filegroup | "default" } can also be specified in a PRIMARY KEY or UNIQUE constraint. These constraints create indexes. If filegroup is specified, the index is stored in the named filegroup. If "default" is specified, or if ON isn't specified at all, the index is stored in the same filegroup as the table. If the PRIMARY KEY or UNIQUE constraint creates a clustered index, the data pages for the table are stored in the same filegroup as the index. If CLUSTERED is specified or the constraint otherwise creates a clustered index, and a partition_scheme is specified that differs from the partition_scheme or filegroup of the table definition, or vice-versa, only the constraint definition will be honored, and the other will be ignored.

Note

In this context, default is not a keyword. It is an identifier for the default filegroup and must be delimited, as in ON "default" or ON [default]. If "default" is specified, the QUOTED_IDENTIFIER option must be ON for the current session. This is the default setting. For more information, see SET QUOTED_IDENTIFIER.

After you create a partitioned table, consider setting the LOCK_ESCALATION option for the table to AUTO. This can improve concurrency by enabling locks to escalate to partition (HoBT) level instead of the table. For more information, see ALTER TABLE.

TEXTIMAGE_ON { filegroup | "default" }

Indicates that the text, ntext, image, xml, varchar(max), nvarchar(max), varbinary(max), and CLR user-defined type columns (including geometry and geography) are stored on the specified filegroup.

TEXTIMAGE_ON isn't allowed if there are no large value columns in the table. TEXTIMAGE_ON can't be specified if partition_scheme is specified. If "default" is specified, or if TEXTIMAGE_ON isn't specified at all, the large value columns are stored in the default filegroup. The storage of any large value column data specified in CREATE TABLE can't be subsequently altered.

Note

varchar(max), nvarchar(max), varbinary(max), xml and large UDT values are stored directly in the data row, up to a limit of 8,000 bytes, and as long as the value can fit the record. If the value does not fit in the record, a pointer is stored in-row and the rest is stored out of row in the LOB storage space. 0 is the default value, which indicates that all values are stored directly in the data row.

TEXTIMAGE_ON only changes the location of the "LOB storage space", it does not affect when data is stored in-row. Use large value types out of row option of sp_tableoption to store the entire LOB value out of the row.

In this context, default is not a keyword. It is an identifier for the default filegroup and must be delimited, as in TEXTIMAGE_ON "default" or TEXTIMAGE_ON [default]. If "default" is specified, the QUOTED_IDENTIFIER option must be ON for the current session. This is the default setting. For more information, see SET QUOTED_IDENTIFIER.

FILESTREAM_ON { partition_scheme_name | filegroup | "default" }

Applies to: SQL Server 2008 R2 (10.50.x) and later. Azure SQL Database and Azure SQL Managed Instance do not support FILESTREAM.

Specifies the filegroup for FILESTREAM data.

If the table contains FILESTREAM data and the table is partitioned, the FILESTREAM_ON clause must be included, and must specify a partition scheme of FILESTREAM filegroups. This partition scheme must use the same partition function and partition columns as the partition scheme for the table; otherwise, an error is raised.

If the table isn't partitioned, the FILESTREAM column can't be partitioned. FILESTREAM data for the table must be stored in a single filegroup. This filegroup is specified in the FILESTREAM_ON clause.

If the table isn't partitioned and the FILESTREAM_ON clause isn't specified, the FILESTREAM filegroup that has the DEFAULT property set is used. If there is no FILESTREAM filegroup, an error is raised.

As with ON and TEXTIMAGE_ON, the value set by using CREATE TABLE for FILESTREAM_ON can't be changed, except in the following cases:

  • A CREATE INDEX statement converts a heap into a clustered index. In this case, a different FILESTREAM filegroup, partition scheme, or NULL can be specified.
  • A DROP INDEX statement converts a clustered index into a heap. In this case, a different FILESTREAM filegroup, partition scheme, or "default" can be specified.

The filegroup in the FILESTREAM_ON <filegroup> clause, or each FILESTREAM filegroup that is named in the partition scheme, must have one file defined for the filegroup. This file must be defined by using a CREATE DATABASE or ALTER DATABASE statement; otherwise, an error is raised.

For related FILESTREAM articles, see Binary Large Object - Blob Data.

[ type_schema_name. ] type_name

Specifies the data type of the column, and the schema to which it belongs. For disk-based tables, use one of the following data types:

  • A system data type
  • An alias type based on a SQL Server system data type. Alias data types are created with the CREATE TYPE statement before they can be used in a table definition. The NULL or NOT NULL assignment for an alias data type can be overridden during the CREATE TABLE statement. However, the length specification can't be changed; the length for an alias data type can't be specified in a CREATE TABLE statement.
  • A CLR user-defined type. CLR user-defined types are created with the CREATE TYPE statement before they can be used in a table definition. To create a column on CLR user-defined type, REFERENCES permission is required on the type.

If type_schema_name isn't specified, the SQL Server Database Engine references type_name in the following order:

  • The SQL Server system data type.
  • The default schema of the current user in the current database.
  • The dbo schema in the current database.

For memory-optimized tables, see Supported Data Types for In-Memory OLTP for a list of supported system types.

  • precision

    The precision for the specified data type. For more information about valid precision values, see Precision, Scale, and Length.

  • scale

    The scale for the specified data type. For more information about valid scale values, see Precision, Scale, and Length.

  • max

    Applies only to the varchar, nvarchar, and varbinary data types for storing 2^31 bytes of character and binary data, and 2^30 bytes of Unicode data.

CONTENT

Specifies that each instance of the xml data type in column_name can contain multiple top-level elements. CONTENT applies only to the xml data type and can be specified only if xml_schema_collection is also specified. If not specified, CONTENT is the default behavior.

DOCUMENT

Specifies that each instance of the xml data type in column_name can contain only one top-level element. DOCUMENT applies only to the xml data type and can be specified only if xml_schema_collection is also specified.

xml_schema_collection

Applies only to the xml data type for associating an XML schema collection with the type. Before typing an xml column to a schema, the schema must first be created in the database by using CREATE XML SCHEMA COLLECTION.

DEFAULT

Specifies the value provided for the column when a value isn't explicitly supplied during an insert. DEFAULT definitions can be applied to any columns except those defined as timestamp, or those with the IDENTITY property. If a default value is specified for a user-defined type column, the type should support an implicit conversion from constant_expression to the user-defined type. DEFAULT definitions are removed when the table is dropped. Only a constant value, such as a character string; a scalar function (either a system, user-defined, or CLR function); or NULL can be used as a default. To maintain compatibility with earlier versions of SQL Server, a constraint name can be assigned to a DEFAULT.

  • constant_expression

    A constant, NULL, or a system function that is used as the default value for the column.

  • memory_optimized_constant_expression

    A constant, NULL, or a system function that is supported in used as the default value for the column. Must be supported in natively compiled stored procedures. For more information about built-in functions in natively compiled stored procedures, see Supported Features for Natively Compiled T-SQL Modules.

IDENTITY

Indicates that the new column is an identity column. When a new row is added to the table, the Database Engine provides a unique, incremental value for the column. Identity columns are typically used with PRIMARY KEY constraints to serve as the unique row identifier for the table. The IDENTITY property can be assigned to tinyint, smallint, int, bigint, decimal(p, 0), or numeric(p, 0) columns. Only one identity column can be created per table. Bound defaults and DEFAULT constraints can't be used with an identity column. Both the seed and increment or neither must be specified. If neither is specified, the default is (1,1).

  • seed

    The value used for the first row loaded into the table.

  • increment

    The incremental value added to the identity value of the previous row loaded.

NOT FOR REPLICATION

In the CREATE TABLE statement, the NOT FOR REPLICATION clause can be specified for the IDENTITY property, FOREIGN KEY constraints, and CHECK constraints. If this clause is specified for the IDENTITY property, values aren't incremented in identity columns when replication agents perform inserts. If this clause is specified for a constraint, the constraint isn't enforced when replication agents perform insert, update, or delete operations.

GENERATED ALWAYS AS { ROW | TRANSACTION_ID | SEQUENCE_NUMBER } { START | END } [ HIDDEN ] [ NOT NULL ]

Applies to: SQL Server 2016 (13.x) and later, Azure SQL Database, and Azure SQL Managed Instance.

Specifies a column used by the system to automatically record information about row versions in the table and its history table (if the table is system versioned and has a history table). Use this argument with the WITH SYSTEM_VERSIONING = ON parameter to create system-versioned tables: temporal or ledger tables. For more information, see updateable ledger tables and temporal tables.

Parameter Required data type Required nullability Description
ROW datetime2 START: NOT NULL
END: NOT NULL
Either the start time for which a row version is valid (START) or the end time for which a row version is valid (END). Use this argument with the PERIOD FOR SYSTEM_TIME argument to create a temporal table.
TRANSACTION_ID bigint START: NOT NULL
END: NULL
Applies to: SQL Server 2022 (16.x) and later, and Azure SQL Database.

The ID of the transaction that creates (START) or invalidates (END) a row version. If the table is a ledger table, the ID references a row in the sys.database_ledger_transactions view
SEQUENCE_NUMBER bigint START: NOT NULL
END: NULL
Applies to: SQL Server 2022 (16.x) and later, and Azure SQL Database.

The sequence number of an operation that creates (START) or deletes (END) a row version. This value is unique within the transaction.

If you attempt to specify a column that doesn't meet the above data type or nullability requirements, the system will throw an error. If you don't explicitly specify nullability, the system will define the column as NULL or NOT NULL per the above requirements.

You can mark one or both period columns with HIDDEN flag to implicitly hide these columns such that SELECT * FROM <table> doesn't return a value for those columns. By default, period columns aren't hidden. In order to be used, hidden columns must be explicitly included in all queries that directly reference the temporal table. To change the HIDDEN attribute for an existing period column, PERIOD must be dropped and recreated with a different hidden flag.

INDEX index_name [ CLUSTERED | NONCLUSTERED ] ( column_name [ ASC | DESC ] [ ,... n ] )

Applies to: SQL Server 2014 (12.x) and later, and Azure SQL Database.

Specifies to create an index on the table. This can be a clustered index, or a nonclustered index. The index will contain the columns listed, and will sort the data in either ascending or descending order.

INDEX index_name CLUSTERED COLUMNSTORE

Applies to: SQL Server 2014 (12.x) and later, and Azure SQL Database.

Specifies to store the entire table in columnar format with a clustered columnstore index. This always includes all columns in the table. The data isn't sorted in alphabetical or numeric order since the rows are organized to gain columnstore compression benefits.

In Azure Synapse Analytics, Analytics Platform System (PDW), and SQL Server 2022 (16.x) and later versions, you can determine the order of the columns for a clustered columnstore index. For more information, see Use an ordered clustered columnstore index for large data warehouse tables.

INDEX index_name [ NONCLUSTERED ] COLUMNSTORE ( column_name [ ,... n ] )

Applies to: SQL Server 2014 (12.x) and later, and Azure SQL Database.

Specifies to create a nonclustered columnstore index on the table. The underlying table can be a rowstore heap or clustered index, or it can be a clustered columnstore index. In all cases, creating a nonclustered columnstore index on a table stores a second copy of the data for the columns in the index.

The nonclustered columnstore index is stored and managed as a clustered columnstore index. It is called a nonclustered columnstore index to because the columns can be limited and it exists as a secondary index on a table.

ON partition_scheme_name ( column_name )

Specifies the partition scheme that defines the filegroups onto which the partitions of a partitioned index will be mapped. The partition scheme must exist within the database by executing either CREATE PARTITION SCHEME or ALTER PARTITION SCHEME. column_name specifies the column against which a partitioned index will be partitioned. This column must match the data type, length, and precision of the argument of the partition function that partition_scheme_name is using. column_name isn't restricted to the columns in the index definition. Any column in the base table can be specified, except when partitioning a UNIQUE index, column_name must be chosen from among those used as the unique key. This restriction allows the Database Engine to verify uniqueness of key values within a single partition only.

Note

When you partition a non-unique, clustered index, the Database Engine by default adds the partitioning column to the list of clustered index keys, if it is not already specified. When partitioning a non-unique, nonclustered index, the Database Engine adds the partitioning column as a non-key (included) column of the index, if it is not already specified.

If partition_scheme_name or filegroup isn't specified and the table is partitioned, the index is placed in the same partition scheme, using the same partitioning column, as the underlying table.

Note

You cannot specify a partitioning scheme on an XML index. If the base table is partitioned, the XML index uses the same partition scheme as the table.

For more information about partitioning indexes, Partitioned Tables and Indexes.

ON filegroup_name

Creates the specified index on the specified filegroup. If no location is specified and the table or view isn't partitioned, the index uses the same filegroup as the underlying table or view. The filegroup must already exist.

ON "default"

Creates the specified index on the default filegroup.

Note

In this context, default is not a keyword. It is an identifier for the default filegroup and must be delimited, as in ON "default" or ON [default]. If "default" is specified, the QUOTED_IDENTIFIER option must be ON for the current session. This is the default setting. For more information, see SET QUOTED_IDENTIFIER.

[ FILESTREAM_ON { filestream_filegroup_name | partition_scheme_name | "NULL" } ]

Applies to: SQL Server 2008 R2 (10.50.x) and later.

Specifies the placement of FILESTREAM data for the table when a clustered index is created. The FILESTREAM_ON clause allows FILESTREAM data to be moved to a different FILESTREAM filegroup or partition scheme.

filestream_filegroup_name is the name of a FILESTREAM filegroup. The filegroup must have one file defined for the filegroup by using a CREATE DATABASE or ALTER DATABASE statement; otherwise, an error is raised.

If the table is partitioned, the FILESTREAM_ON clause must be included, and must specify a partition scheme of FILESTREAM filegroups that uses the same partition function and partition columns as the partition scheme for the table. Otherwise, an error is raised.

If the table isn't partitioned, the FILESTREAM column can't be partitioned. FILESTREAM data for the table must be stored in a single filegroup that is specified in the FILESTREAM_ON clause.

FILESTREAM_ON NULL can be specified in a CREATE INDEX statement if a clustered index is being created and the table doesn't contain a FILESTREAM column.

For more information, see FILESTREAM.

ROWGUIDCOL

Indicates that the new column is a row GUID column. Only one uniqueidentifier column per table can be designated as the ROWGUIDCOL column. Applying the ROWGUIDCOL property enables the column to be referenced using $ROWGUID. The ROWGUIDCOL property can be assigned only to a uniqueidentifier column. User-defined data type columns can't be designated with ROWGUIDCOL.

The ROWGUIDCOL property doesn't enforce uniqueness of the values stored in the column. ROWGUIDCOL also doesn't automatically generate values for new rows inserted into the table. To generate unique values for each column, either use the NEWID or NEWSEQUENTIALID function on INSERT statements or use these functions as the default for the column.

ENCRYPTED WITH

Specifies encrypting columns by using the Always Encrypted feature.

  • COLUMN_ENCRYPTION_KEY = key_name

    Specifies the column encryption key. For more information, see CREATE COLUMN ENCRYPTION KEY.

  • ENCRYPTION_TYPE = { DETERMINISTIC | RANDOMIZED }

    Deterministic encryption uses a method that always generates the same encrypted value for any given plain text value. Using deterministic encryption allows searching using equality comparison, grouping, and joining tables using equality joins based on encrypted values, but can also allow unauthorized users to guess information about encrypted values by examining patterns in the encrypted column. Joining two tables on columns encrypted deterministically is only possible if both columns are encrypted using the same column encryption key. Deterministic encryption must use a column collation with a binary2 sort order for character columns.

    Randomized encryption uses a method that encrypts data in a less predictable manner. Randomized encryption is more secure, but it prevents any computations and indexing on encrypted columns, unless your SQL Server instance supports Always Encrypted with secure enclaves. See Always Encrypted with secure enclaves for details.

    If you are using Always Encrypted (without secure enclaves), use deterministic encryption for columns that will be searched with parameters or grouping parameters, for example a government ID number. Use randomized encryption, for data such as a credit card number, which isn't grouped with other records or used to join tables, and which isn't searched for because you use other columns (such as a transaction number) to find the row that contains the encrypted column of interest.

    If you are using Always Encrypted with secure enclaves, randomized encryption is a recommended encryption type.

    Columns must be of a qualifying data type.

  • ALGORITHM

    Applies to: SQL Server 2016 (13.x) and later.

    Must be 'AEAD_AES_256_CBC_HMAC_SHA_256'.

    For more information including feature constraints, see Always Encrypted.

SPARSE

Indicates that the column is a sparse column. The storage of sparse columns is optimized for null values. Sparse columns can't be designated as NOT NULL. For additional restrictions and more information about sparse columns, see Use Sparse Columns.

MASKED WITH ( FUNCTION = 'mask_function' )

Applies to: SQL Server 2016 (13.x) and later.

Specifies a dynamic data mask. mask_function is the name of the masking function with the appropriate parameters. Four functions are available:

  • default()
  • email()
  • partial()
  • random()

Requires ALTER ANY MASK permission.

For function parameters, see Dynamic Data Masking.

FILESTREAM

Applies to: SQL Server 2008 R2 (10.50.x) and later.

Valid only for varbinary(max) columns. Specifies FILESTREAM storage for the varbinary(max) BLOB data.

The table must also have a column of the uniqueidentifier data type that has the ROWGUIDCOL attribute. This column must not allow null values and must have either a UNIQUE or PRIMARY KEY single-column constraint. The GUID value for the column must be supplied either by an application when inserting data, or by a DEFAULT constraint that uses the NEWID () function.

The ROWGUIDCOL column can't be dropped and the related constraints can't be changed while there is a FILESTREAM column defined for the table. The ROWGUIDCOL column can be dropped only after the last FILESTREAM column is dropped.

When the FILESTREAM storage attribute is specified for a column, all values for that column are stored in a FILESTREAM data container on the file system.

COLLATE collation_name

Specifies the collation for the column. Collation name can be either a Windows collation name or a SQL collation name. collation_name is applicable only for columns of the char, varchar, text, nchar, nvarchar, and ntext data types. If not specified, the column is assigned either the collation of the user-defined data type, if the column is of a user-defined data type, or the default collation of the database.

For more information about the Windows and SQL collation names, see Windows Collation Name and SQL Collation Name.

For more information, see COLLATE.

CONSTRAINT

An optional keyword that indicates the start of the definition of a PRIMARY KEY, NOT NULL, UNIQUE, FOREIGN KEY, or CHECK constraint.

  • constraint_name

    The name of a constraint. Constraint names must be unique within the schema to which the table belongs.

  • NULL | NOT NULL

    Determine whether null values are allowed in the column. NULL isn't strictly a constraint but can be specified just like NOT NULL. NOT NULL can be specified for computed columns only if PERSISTED is also specified.

  • PRIMARY KEY

    A constraint that enforces entity integrity for a specified column or columns through a unique index. Only one PRIMARY KEY constraint can be created per table.

  • UNIQUE

    A constraint that provides entity integrity for a specified column or columns through a unique index. A table can have multiple UNIQUE constraints.

  • CLUSTERED | NONCLUSTERED

    Indicates that a clustered or a nonclustered index is created for the PRIMARY KEY or UNIQUE constraint. PRIMARY KEY constraints default to CLUSTERED, and UNIQUE constraints default to NONCLUSTERED.

    In a CREATE TABLE statement, CLUSTERED can be specified for only one constraint. If CLUSTERED is specified for a UNIQUE constraint and a PRIMARY KEY constraint is also specified, the PRIMARY KEY defaults to NONCLUSTERED.

  • FOREIGN KEY REFERENCES

    A constraint that provides referential integrity for the data in the column or columns. FOREIGN KEY constraints require that each value in the column exists in the corresponding referenced column or columns in the referenced table. FOREIGN KEY constraints can reference only columns that are PRIMARY KEY or UNIQUE constraints in the referenced table or columns referenced in a UNIQUE INDEX on the referenced table. Foreign keys on computed columns must also be marked PERSISTED.

  • [ [ schema_name. ] referenced_table_name ]

    The name of the table referenced by the FOREIGN KEY constraint, and the schema to which it belongs.

  • ( ref_column [ ,... n ] )

    A column, or list of columns, from the table referenced by the FOREIGN KEY constraint.

  • ON DELETE { NO ACTION | CASCADE | SET NULL | SET DEFAULT }

    Specifies what action happens to rows in the table created, if those rows have a referential relationship and the referenced row is deleted from the parent table. The default is NO ACTION.

  • NO ACTION

    The Database Engine raises an error and the delete action on the row in the parent table is rolled back.

  • CASCADE

    Corresponding rows are deleted from the referencing table if that row is deleted from the parent table.

  • SET NULL

    All the values that make up the foreign key are set to NULL if the corresponding row in the parent table is deleted. For this constraint to execute, the foreign key columns must be nullable.

  • SET DEFAULT

    All the values that make up the foreign key are set to their default values when the corresponding row in the parent table is deleted. For this constraint to execute, all foreign key columns must have default definitions. If a column is nullable, and there is no explicit default value set, NULL becomes the implicit default value of the column.

    Don't specify CASCADE if the table will be included in a merge publication that uses logical records. For more information about logical records, see Group Changes to Related Rows with Logical Records.

    ON DELETE CASCADE can't be defined if an INSTEAD OF trigger ON DELETE already exists on the table.

    For example, in the AdventureWorks2022 database, the ProductVendor table has a referential relationship with the Vendor table. The ProductVendor.BusinessEntityID foreign key references the Vendor.BusinessEntityID primary key.

    If a DELETE statement is executed on a row in the Vendor table, and an ON DELETE CASCADE action is specified for ProductVendor.BusinessEntityID, the Database Engine checks for one or more dependent rows in the ProductVendor table. If any exist, the dependent rows in the ProductVendor table are deleted, and also the row referenced in the Vendor table.

    Conversely, if NO ACTION is specified, the Database Engine raises an error and rolls back the delete action on the Vendor row if there is at least one row in the ProductVendor table that references it.

  • ON UPDATE { NO ACTION | CASCADE | SET NULL | SET DEFAULT }

    Specifies what action happens to rows in the table altered when those rows have a referential relationship and the referenced row is updated in the parent table. The default is NO ACTION.

  • NO ACTION

    The Database Engine raises an error, and the update action on the row in the parent table is rolled back.

  • CASCADE

    Corresponding rows are updated in the referencing table when that row is updated in the parent table.

  • SET NULL

    All the values that make up the foreign key are set to NULL when the corresponding row in the parent table is updated. For this constraint to execute, the foreign key columns must be nullable.

  • SET DEFAULT

    All the values that make up the foreign key are set to their default values when the corresponding row in the parent table is updated. For this constraint to execute, all foreign key columns must have default definitions. If a column is nullable, and there is no explicit default value set, NULL becomes the implicit default value of the column.

    Don't specify CASCADE if the table will be included in a merge publication that uses logical records. For more information about logical records, see Group Changes to Related Rows with Logical Records.

    ON UPDATE CASCADE, SET NULL, or SET DEFAULT can't be defined if an INSTEAD OF trigger ON UPDATE already exists on the table that is being altered.

    For example, in the AdventureWorks2022 database, the ProductVendor table has a referential relationship with the Vendor table: ProductVendor.BusinessEntity foreign key references the Vendor.BusinessEntityID primary key.

    If an UPDATE statement is executed on a row in the Vendor table, and an ON UPDATE CASCADE action is specified for ProductVendor.BusinessEntityID, the Database Engine checks for one or more dependent rows in the ProductVendor table. If any exist, the dependent rows in the ProductVendor table are updated, and also the row referenced in the Vendor table.

    Conversely, if NO ACTION is specified, the Database Engine raises an error and rolls back the update action on the Vendor row if there is at least one row in the ProductVendor table that references it.

  • CHECK

    A constraint that enforces domain integrity by limiting the possible values that can be entered into a column or columns. CHECK constraints on computed columns must also be marked PERSISTED.

  • logical_expression

    A logical expression that returns TRUE or FALSE. Alias data types can't be part of the expression.

  • column_name

    A column or list of columns, in parentheses, used in table constraints to indicate the columns used in the constraint definition.

  • [ ASC | DESC ]

    Specifies the order in which the column or columns participating in table constraints are sorted. The default is ASC.

  • partition_scheme_name

    The name of the partition scheme that defines the filegroups onto which the partitions of a partitioned table will be mapped. The partition scheme must exist within the database.

  • [ partition_column_name. ]

    Specifies the column against which a partitioned table will be partitioned. The column must match that specified in the partition function that partition_scheme_name is using in terms of data type, length, and precision. A computed column that participates in a partition function must be explicitly marked PERSISTED.

    Important

    We recommend that you specify NOT NULL on the partitioning column of partitioned tables, and also nonpartitioned tables that are sources or targets of ALTER TABLE...SWITCH operations. Doing this makes sure that any CHECK constraints on partitioning columns do not have to check for null values.

  • WITH FILLFACTOR = fillfactor

    Specifies how full the Database Engine should make each index page that is used to store the index data. User-specified fillfactor values can be from 1 through 100. If a value isn't specified, the default is 0. Fill factor values 0 and 100 are the same in all respects.

    Important

    Documenting WITH FILLFACTOR = fillfactor as the only index option that applies to PRIMARY KEY or UNIQUE constraints is maintained for backward compatibility, but will not be documented in this manner in future releases.

column_set_name XML COLUMN_SET FOR ALL_SPARSE_COLUMNS

The name of the column set. A column set is an untyped XML representation that combines all of the sparse columns of a table into a structured output. For more information about column sets, see Use Column Sets.

PERIOD FOR SYSTEM_TIME ( system_start_time_column_name , system_end_time_column_name )

Applies to: SQL Server 2016 (13.x) and later, and Azure SQL Database.

Specifies the names of the columns that the system will use to record the period for which a record is valid. Use this argument with the GENERATED ALWAYS AS ROW { START | END } and WITH SYSTEM_VERSIONING = ON arguments to create a temporal table. For more information, see Temporal Tables.

COMPRESSION_DELAY

Applies to: SQL Server 2016 (13.x) and later, and Azure SQL Database.

For a memory-optimized, delay specifies the minimum number of minutes a row must remain in the table, unchanged, before it is eligible for compression into the columnstore index. SQL Server selects specific rows to compress according to their last update time. For example, if rows are changing frequently during a two-hour period of time, you could set COMPRESSION_DELAY = 120 Minutes to ensure updates are completed before SQL Server compresses the row.

For a disk-based table, delay specifies the minimum number of minutes a delta rowgroup in the CLOSED state must remain in the delta rowgroup before SQL Server can compress it into the compressed rowgroup. Since disk-based tables don't track insert and update times on individual rows, SQL Server applies the delay to delta rowgroups in the CLOSED state.

The default is 0 minutes.

For recommendations on when to use COMPRESSION_DELAY, see Get started with Columnstore for real time operational analytics

<table_option> ::=

Specifies one or more table options.

DATA_COMPRESSION

Specifies the data compression option for the specified table, partition number, or range of partitions. The options are as follows:

  • NONE

    Table or specified partitions aren't compressed.

  • ROW

    Table or specified partitions are compressed by using row compression.

  • PAGE

    Table or specified partitions are compressed by using page compression.

  • COLUMNSTORE

    Applies to: SQL Server 2016 (13.x) and later, and Azure SQL Database.

    Applies only to columnstore indexes, including both nonclustered columnstore and clustered columnstore indexes. COLUMNSTORE specifies to compress with the most performant columnstore compression. This is the typical choice.

  • COLUMNSTORE_ARCHIVE

    Applies to: SQL Server 2016 (13.x) and later, and Azure SQL Database.

    Applies only to columnstore indexes, including both nonclustered columnstore and clustered columnstore indexes. COLUMNSTORE_ARCHIVE will further compress the table or partition to a smaller size. This can be used for archival, or for other situations that require a smaller storage size and can afford more time for storage and retrieval.

For more information, see Data Compression.

XML_COMPRESSION

Applies to: SQL Server 2022 (16.x) and later versions, Azure SQL Database, and Azure SQL Managed Instance.

Specifies the XML compression option for any xml data type columns in the table. The options are as follows:

  • ON

    Columns using the xml data type are compressed.

  • OFF

    Columns using the xml data type aren't compressed.

ON PARTITIONS ( { <partition_number_expression> | [ ,... n ] )

Specifies the partitions to which the DATA_COMPRESSION or XML_COMPRESSION settings apply. If the table isn't partitioned, the ON PARTITIONS argument will generate an error. If the ON PARTITIONS clause isn't provided, the DATA_COMPRESSION option will apply to all partitions of a partitioned table.

partition_number_expression can be specified in the following ways:

  • Provide the partition number of a partition, for example: ON PARTITIONS (2)
  • Provide the partition numbers for several individual partitions separated by commas, for example: ON PARTITIONS (1, 5)
  • Provide both ranges and individual partitions, for example: ON PARTITIONS (2, 4, 6 TO 8)

<range> can be specified as partition numbers separated by the word TO, for example: ON PARTITIONS (6 TO 8).

To set different types of data compression for different partitions, specify the DATA_COMPRESSION option more than once, for example:

WITH
(
    DATA_COMPRESSION = NONE ON PARTITIONS (1),
    DATA_COMPRESSION = ROW ON PARTITIONS (2, 4, 6 TO 8),
    DATA_COMPRESSION = PAGE ON PARTITIONS (3, 5)
);

You can also specify the XML_COMPRESSION option more than once, for example:

WITH
(
    XML_COMPRESSION = OFF ON PARTITIONS (1),
    XML_COMPRESSION = ON ON PARTITIONS (2, 4, 6 TO 8),
    XML_COMPRESSION = OFF ON PARTITIONS (3, 5)
);

<index_option> ::=

Specifies one or more index options. For a complete description of these options, see CREATE INDEX.

PAD_INDEX = { ON | OFF }

When ON, the percentage of free space specified by FILLFACTOR is applied to the intermediate level pages of the index. When OFF or a FILLFACTOR value it not specified, the intermediate level pages are filled to near capacity leaving enough space for at least one row of the maximum size the index can have, considering the set of keys on the intermediate pages. The default is OFF.

FILLFACTOR = fillfactor

Specifies a percentage that indicates how full the Database Engine should make the leaf level of each index page during index creation or alteration. fillfactor must be an integer value from 1 to 100. The default is 0. Fill factor values 0 and 100 are the same in all respects.

IGNORE_DUP_KEY = { ON | OFF }

Specifies the error response when an insert operation attempts to insert duplicate key values into a unique index. The IGNORE_DUP_KEY option applies only to insert operations after the index is created or rebuilt. The option has no effect when executing CREATE INDEX, ALTER INDEX, or UPDATE. The default is OFF.

  • ON

    A warning message will occur when duplicate key values are inserted into a unique index. Only the rows violating the uniqueness constraint will fail.

  • OFF

    An error message will occur when duplicate key values are inserted into a unique index. The entire INSERT operation will be rolled back.

IGNORE_DUP_KEY can't be set to ON for indexes created on a view, non-unique indexes, XML indexes, spatial indexes, and filtered indexes.

To view IGNORE_DUP_KEY, use sys.indexes.

In backward compatible syntax, WITH IGNORE_DUP_KEY is equivalent to WITH IGNORE_DUP_KEY = ON.

STATISTICS_NORECOMPUTE = { ON | OFF }

When ON, out-of-date index statistics aren't automatically recomputed. When OFF, automatic statistics updating are enabled. The default is OFF.

ALLOW_ROW_LOCKS = { ON | OFF }

When ON, row locks are allowed when you access the index. The Database Engine determines when row locks are used. When OFF, row locks aren't used. The default is ON.

ALLOW_PAGE_LOCKS = { ON | OFF }

When ON, page locks are allowed when you access the index. The Database Engine determines when page locks are used. When OFF, page locks aren't used. The default is ON.

OPTIMIZE_FOR_SEQUENTIAL_KEY = { ON | OFF }

Applies to: SQL Server 2019 (15.x) and later, Azure SQL Database, and Azure SQL Managed Instance.

Specifies whether or not to optimize for last-page insert contention. The default is OFF. See the Sequential Keys section of the CREATE INDEX page for more information.

FILETABLE_DIRECTORY = directory_name

Applies to: SQL Server 2012 (11.x) and later.

Specifies the windows-compatible FileTable directory name. This name should be unique among all the FileTable directory names in the database. Uniqueness comparison is case-insensitive, regardless of collation settings. If this value isn't specified, the name of the FileTable is used.

FILETABLE_COLLATE_FILENAME = { collation_name | database_default }

Applies to: SQL Server 2012 (11.x) and later. Azure SQL Database and Azure SQL Managed Instance do not support FILETABLE.

Specifies the name of the collation to be applied to the Name column in the FileTable. The collation must be case-insensitive to comply with Windows operating system file naming semantics. If this value isn't specified, the database default collation is used. If the database default collation is case-sensitive, an error is raised, and the CREATE TABLE operation fails.

  • collation_name

    The name of a case-insensitive collation.

  • database_default

    Specifies that the default collation for the database should be used. This collation must be case-insensitive.

FILETABLE_PRIMARY_KEY_CONSTRAINT_NAME = constraint_name

Applies to: SQL Server 2012 (11.x) and later. Azure SQL Database and Azure SQL Managed Instance do not support FILETABLE.

Specifies the name to be used for the primary key constraint that is automatically created on the FileTable. If this value isn't specified, the system generates a name for the constraint.

FILETABLE_STREAMID_UNIQUE_CONSTRAINT_NAME = constraint_name

Applies to: SQL Server 2012 (11.x) and later. Azure SQL Database and Azure SQL Managed Instance do not support FILETABLE.

Specifies the name to be used for the unique constraint that is automatically created on the stream_id column in the FileTable. If this value isn't specified, the system generates a name for the constraint.

FILETABLE_FULLPATH_UNIQUE_CONSTRAINT_NAME = constraint_name

Applies to: SQL Server 2012 (11.x) and later. Azure SQL Database and Azure SQL Managed Instance do not support FILETABLE.

Specifies the name to be used for the unique constraint that is automatically created on the parent_path_locator and name columns in the FileTable. If this value isn't specified, the system generates a name for the constraint.

SYSTEM_VERSIONING = ON [ ( HISTORY_TABLE = schema_name.history_table_name [ , DATA_CONSISTENCY_CHECK = { ON | OFF } ] ) ]

Applies to: SQL Server 2016 (13.x) and later, Azure SQL Database, and Azure SQL Managed Instance.

Enables system versioning of the table if the datatype, nullability constraint, and primary key constraint requirements are met. The system will record the history of each record in the system-versioned table in a separate history table. If the HISTORY_TABLE argument isn't used, the name of this history table will be MSSQL_TemporalHistoryFor<primary_table_object_id>. If the name of a history table is specified during history table creation, you must specify the schema and table name.

If the history table doesn't exist, the system generates a new history table matching the schema of the current table in the same filegroup as the current table, creating a link between the two tables and enables the system to record the history of each record in the current table in the history table. By default, the history table is PAGE compressed.

If the HISTORY_TABLE argument is used to create a link to and use an existing history table, the link is created between the current table and the specified table. If current table is partitioned, the history table is created on default file group because partitioning configuration isn't replicated automatically from the current table to the history table. When creating a link to an existing history table, you can choose to perform a data consistency check. This data consistency check ensures that existing records don't overlap. Performing the data consistency check is the default.

Use this argument with the PERIOD FOR SYSTEM_TIME and GENERATED ALWAYS AS ROW { START | END } arguments to enable system versioning on a table. For more information, see Temporal Tables. Use this argument with the WITH LEDGER = ON argument to create an updatable ledger table. Using existing history tables with ledger tables isn't allowed.

REMOTE_DATA_ARCHIVE = { ON [ ( table_stretch_options [ ,... n ] ) ] | OFF ( MIGRATION_STATE = PAUSED ) }

Applies to: SQL Server 2016 (13.x) and later.

Creates the new table with Stretch Database enabled or disabled. For more info, see Stretch Database.

Important

Stretch Database is deprecated in SQL Server 2022 (16.x) and Azure SQL Database. This feature will be removed in a future version of the Database Engine. Avoid using this feature in new development work, and plan to modify applications that currently use this feature.

Enabling Stretch Database for a table

When you enable Stretch for a table by specifying ON, you can optionally specify MIGRATION_STATE = OUTBOUND to begin migrating data immediately, or MIGRATION_STATE = PAUSED to postpone data migration. The default value is MIGRATION_STATE = OUTBOUND. For more info about enabling Stretch for a table, see Enable Stretch Database for a table.

Prerequisites. Before you enable Stretch for a table, you have to enable Stretch on the server and on the database. For more info, see Enable Stretch Database for a database.

Permissions. Enabling Stretch for a database or a table requires db_owner permissions. Enabling Stretch for a table also requires ALTER permissions on the table.

[ FILTER_PREDICATE = { NULL | predicate } ]

Applies to: SQL Server 2016 (13.x) and later.

Optionally specifies a filter predicate to select rows to migrate from a table that contains both historical and current data. The predicate must call a deterministic inline table-valued function. For more info, see Enable Stretch Database for a table and Select rows to migrate by using a filter function.

Important

If you provide a filter predicate that performs poorly, data migration also performs poorly. Stretch Database applies the filter predicate to the table by using the CROSS APPLY operator.

If you don't specify a filter predicate, the entire table is migrated.

When you specify a filter predicate, you also have to specify MIGRATION_STATE.

MIGRATION_STATE = { OUTBOUND | INBOUND | PAUSED }

Applies to: SQL Server 2016 (13.x) and later, Azure SQL Database, and Azure SQL Managed Instance.

[ DATA_DELETION = ON { ( FILTER_COLUMN = column_name, RETENTION_PERIOD = { INFINITE | number { DAY | DAYS | WEEK | WEEKS | MONTH | MONTHS | YEAR | YEARS } ) } ]

Applies to: Azure SQL Edge only

Enables retention policy based cleanup of old or aged data from tables within a database. For more information, see Enable and Disable Data Retention. The following parameters must be specified for data retention to be enabled.

  • FILTER_COLUMN = { column_name }

    Specifies the column that should be used to determine if the rows in the table are obsolete or not. The following data types are allowed for the filter column.

    • date
    • datetime
    • datetime2
    • smalldatetime
    • datetimeoffset
  • RETENTION_PERIOD = { INFINITE | number {DAY | DAYS | WEEK | WEEKS | MONTH | MONTHS | YEAR | YEARS }}

    Specifies the retention period policy for the table. The retention period is specified as a combination of a positive integer value and the date part unit.

MEMORY_OPTIMIZED

Applies to: SQL Server 2014 (12.x) and later, Azure SQL Database, and Azure SQL Managed Instance. Azure SQL Managed Instance does not support memory optimized tables in General Purpose tier.

The value ON indicates that the table is memory optimized. Memory-optimized tables are part of the In-Memory OLTP feature, which is used to optimize the performance of transaction processing. To get started with In-Memory OLTP see Quickstart 1: In-Memory OLTP Technologies for Faster Transact-SQL Performance. For more in-depth information about memory-optimized tables, see Memory-Optimized Tables.

The default value OFF indicates that the table is disk-based.

DURABILITY

Applies to: SQL Server 2014 (12.x) and later, Azure SQL Database, and Azure SQL Managed Instance.

The value of SCHEMA_AND_DATA indicates that the table is durable, meaning that changes are persisted on disk and survive restart or failover. SCHEMA_AND_DATA is the default value.

The value of SCHEMA_ONLY indicates that the table is non-durable. The table schema is persisted but any data updates aren't persisted upon a restart or failover of the database. DURABILITY = SCHEMA_ONLY is only allowed with MEMORY_OPTIMIZED = ON.

Warning

When a table is created with DURABILITY = SCHEMA_ONLY, and READ_COMMITTED_SNAPSHOT is subsequently changed using ALTER DATABASE, data in the table will be lost.

BUCKET_COUNT

Applies to: SQL Server 2014 (12.x) and later, Azure SQL Database, and Azure SQL Managed Instance.

Indicates the number of buckets that should be created in the hash index. The maximum value for BUCKET_COUNT in hash indexes is 1,073,741,824. For more information about bucket counts, see Indexes for Memory-Optimized Tables.

Bucket_count is a required argument.

INDEX

Applies to: SQL Server 2014 (12.x) and later, Azure SQL Database, and Azure SQL Managed Instance.

Column and table indexes can be specified as part of the CREATE TABLE statement. For details about adding and removing indexes on memory-optimized tables, see Altering Memory-Optimized Tables

  • HASH

    Applies to: SQL Server 2014 (12.x) and later, Azure SQL Database, and Azure SQL Managed Instance.

    Indicates that a HASH index is created.

    Hash indexes are supported only on memory-optimized tables.

LEDGER = ON ( <ledger_option> [ ,... n ] ) | OFF

Applies to: SQL Server 2022 (16.x), Azure SQL Database, and Azure SQL Managed Instance.

Note

If the statement creates a ledger table, the ENABLE LEDGER permission is required.

Indicates whether the table being created is a ledger table (ON) or not (OFF). The default is OFF. If the APPEND_ONLY = ON option is specified, the system creates an append-only ledger table allowing only inserting new rows. Otherwise, the system creates an updatable ledger table. An updatable ledger table also requires the SYSTEM_VERSIONING = ON argument. An updatable ledger table must also be a system-versioned table. However, an updatable ledger table doesn't have to be a temporal table (it doesn't require the PERIOD FOR SYSTEM_TIME parameter). If the history table is specified with LEDGER = ON and SYSTEM_VERSIONING = ON, it must not reference an existing table.

A ledger database (a database created with the LEDGER = ON option) only allows the creation of ledger tables. Attempts to create a table with LEDGER = OFF will raise an error. Each new table by default is created as an updatable ledger table, even if you don't specify LEDGER = ON, and will be created with default values for all other parameters.

An updatable ledger table must contain four GENERATED ALWAYS columns, exactly one column defined with each of the following arguments:

  • GENERATED ALWAYS AS TRANSACTION_ID START
  • GENERATED ALWAYS AS TRANSACTION_ID END
  • GENERATED ALWAYS AS SEQUENCE_NUMBER START
  • GENERATED ALWAYS AS SEQUENCE_NUMBER END

An append-only ledger table must contain exactly one column defined with each of the following arguments:

  • GENERATED ALWAYS AS TRANSACTION_ID START
  • GENERATED ALWAYS AS SEQUENCE_NUMBER START

If any of the required generated always columns isn't defined in the CREATE TABLE statement and the statement includes LEDGER = ON, the system will automatically attempt to add the column using an applicable column definition from the below list. If there is a name conflict with an already defined column, the system will raise an error.

[ledger_start_transaction_id] BIGINT GENERATED ALWAYS AS TRANSACTION_ID START HIDDEN NOT NULL
[ledger_end_transaction_id] BIGINT GENERATED ALWAYS AS TRANSACTION_ID END HIDDEN NULL
[ledger_start_sequence_number] BIGINT GENERATED ALWAYS AS SEQUENCE_NUMBER START HIDDEN NOT NULL
[ledger_end_sequence_number] BIGINT GENERATED ALWAYS AS SEQUENCE_NUMBER END HIDDEN NULL

The <ledger_view_option> specifies the schema and the name of the ledger view the system automatically creates and links to the table. If the option isn't specified, the system generates the ledger view name by appending _Ledger to the name of the table being created (database_name.schema_name.table_name). If a view with the specified or generated name exists, the system will raise an error. If the table is an updatable ledger table, the ledger view is created as a union on the table and its history table.

Each row in the ledger view represents either the creation or deletion of a row version in the ledger table. The ledger view contains all columns of the ledger table, except the generated always columns listed above. The ledger view also contains the following additional columns:

Column name Data type Description
Specified using the TRANSACTION_ID_COLUMN_NAME option. ledger_transaction_id if not specified. bigint The ID of the transaction that created or deleted a row version.
Specified using the SEQUENCE_NUMBER_COLUMN_NAME option. ledger_sequence_number if not specified. bigint The sequence number of a row-level operation within the transaction on the table.
Specified using the OPERATION_TYPE_COLUMN_NAME option. ledger_operation_type if not specified. tinyint Contains 1 (INSERT) or 2 (DELETE). Inserting a row into the ledger table produces a new row in the ledger view containing 1 in this column. Deleting a row from the ledger table produces a new row in the ledger view containing 2 in this column. Updating a row in the ledger table produces two new rows in the ledger view. One row contains 2 (DELETE) and the other row contains 1 (INSERT) in this column.
Specified using the OPERATION_TYPE_DESC_COLUMN_NAME option. ledger_operation_type_desc if not specified. nvarchar(128) Contains INSERT or DELETE. See above for details.

Transactions that include creating ledger table are captured in sys.database_ledger_transactions.

<ledger_option> ::=

Specifies a ledger option.

[ LEDGER_VIEW = schema_name.ledger_view_name [ ( <ledger_view_option> [ ,... n ] ) ]

Specifies the name of the ledger view and the names of additional columns the system adds to the ledger view.

[ APPEND_ONLY = ON | OFF ]

Specifies whether the ledger table being created is append-only or updatable. The default is OFF.

<ledger_view_option> ::=

Specifies one or more ledger view options. Each of the ledger view option specifies a name of a column, the system will add to the view, in addition to the columns defined in the ledger table.

[ TRANSACTION_ID_COLUMN_NAME = transaction_id_column_name ]

Specifies the name of the column storing the ID of the transaction that created or deleted a row version. The default column name is ledger_transaction_id.

[ SEQUENCE_NUMBER_COLUMN_NAME = sequence_number_column_name ]

Specifies the name of the columns storing the sequence number of a row-level operation within the transaction on the table. The default column name is ledger_sequence_number.

[ OPERATION_TYPE_COLUMN_NAME = operation_type_id column_name ]

Specifies the name of the columns storing the operation type ID. The default column name is ledger_operation_type.

[ OPERATION_TYPE_DESC_COLUMN_NAME = operation_type_desc_column_name ]

Specifies the name of the columns storing the operation type description. The default column name is ledger_operation_type_desc.

Remarks

For information about the number of allowed tables, columns, constraints and indexes, see Maximum Capacity Specifications for SQL Server.

Space is generally allocated to tables and indexes in increments of one extent at a time. When the SET MIXED_PAGE_ALLOCATION option of ALTER DATABASE is set to TRUE, or always prior to SQL Server 2016 (13.x), when a table or index is created, it is allocated pages from mixed extents until it has enough pages to fill a uniform extent. After it has enough pages to fill a uniform extent, another extent is allocated every time the currently allocated extents become full. For a report about the amount of space allocated and used by a table, execute sp_spaceused.

The Database Engine doesn't enforce an order in which DEFAULT, IDENTITY, ROWGUIDCOL, or column constraints are specified in a column definition.

When a table is created, the QUOTED IDENTIFIER option is always stored as ON in the metadata for the table, even if the option is set to OFF when the table is created.

In SQL database in Microsoft Fabric, some table features can be created but will not be mirrored into the Fabric OneLake. For more information, see Limitations of Fabric SQL database mirroring.

Temporary tables

You can create local and global temporary tables. Local temporary tables are visible only in the current session, and global temporary tables are visible to all sessions. Temporary tables can't be partitioned.

Prefix local temporary table names with single number sign (#table_name), and prefix global temporary table names with a double number sign (##table_name).

Transact-SQL statements reference the temporary table by using the value specified for table_name in the CREATE TABLE statement, for example:

CREATE TABLE #MyTempTable (
    col1 INT PRIMARY KEY
);

INSERT INTO #MyTempTable
VALUES (1);

If more than one temporary table is created inside a single stored procedure or batch, they must have different names.

If you include a schema_name when you create or access a temporary table, it is ignored. All temporary tables are created in the dbo schema.

If a local temporary table is created in a stored procedure or application that can be executed at the same time by several sessions, the Database Engine must be able to distinguish the tables created by the different sessions. The Database Engine does this by internally appending a numeric suffix to each local temporary table name. The full name of a temporary table as stored in the sys.sysobjects table in tempdb is made up of the table name specified in the CREATE TABLE statement and the system-generated numeric suffix. To allow for the suffix, table_name specified for a local temporary name can't exceed 116 characters.

Temporary tables are automatically dropped when they go out of scope, unless explicitly dropped by using DROP TABLE:

  • A local temporary table created in a stored procedure is dropped automatically when the stored procedure is finished. The table can be referenced by any nested stored procedures executed by the stored procedure that created the table. The table can't be referenced by the process that called the stored procedure that created the table.
  • All other local temporary tables are dropped automatically at the end of the current session.
  • Global temporary tables are automatically dropped when the session that created the table ends and all other tasks have stopped referencing them. The association between a task and a table is maintained only for the life of a single Transact-SQL statement. This means that a global temporary table is dropped at the completion of the last Transact-SQL statement that was actively referencing the table when the creating session ended.

A local temporary table created within a stored procedure or trigger can have the same name as a temporary table that was created before the stored procedure or trigger is called. However, if a query references a temporary table and two temporary tables with the same name exist at that time, it isn't defined which table the query is resolved against. Nested stored procedures can also create temporary tables with the same name as a temporary table that was created by the stored procedure that called it. However, for modifications to resolve to the table that was created in the nested procedure, the table must have the same structure, with the same column names, as the table created in the calling procedure. This is shown in the following example.

CREATE PROCEDURE dbo.Test2
AS
    CREATE TABLE #t (x INT PRIMARY KEY);
    INSERT INTO #t VALUES (2);
    SELECT Test2Col = x FROM #t;
GO

CREATE PROCEDURE dbo.Test1
AS
    CREATE TABLE #t (x INT PRIMARY KEY);
    INSERT INTO #t VALUES (1);
    SELECT Test1Col = x FROM #t;
    EXEC Test2;
GO

CREATE TABLE #t(x INT PRIMARY KEY);
INSERT INTO #t VALUES (99);
GO

EXEC Test1;
GO

Here's the result set.

(1 row(s) affected)
Test1Col
-----------
1

(1 row(s) affected)
 Test2Col
 -----------
 2

When you create local or global temporary tables, the CREATE TABLE syntax supports constraint definitions except for FOREIGN KEY constraints. If a FOREIGN KEY constraint is specified in a temporary table, the statement returns a warning message that states the constraint was skipped. The table is still created without the FOREIGN KEY constraints. Temporary tables can't be referenced in FOREIGN KEY constraints.

If a temporary table is created with a named constraint and the temporary table is created within the scope of a user-defined transaction, only one user at a time can execute the statement that creates the temp table. For example, if a stored procedure creates a temporary table with a named primary key constraint, the stored procedure can't be executed simultaneously by multiple users.

Database scoped global temporary tables (Azure SQL Database)

Global temporary tables for SQL Server (initiated with ## table name) are stored in tempdb and shared among all users' sessions across the whole SQL Server instance. For information on SQL table types, see the above section on Create Tables.

Azure SQL Database supports global temporary tables that are also stored in tempdb and scoped to the database level. This means that global temporary tables are shared for all users' sessions within the same Azure SQL Database. User sessions from other databases can't access global temporary tables.

Global temporary tables for Azure SQL Database follow the same syntax and semantics that SQL Server uses for temporary tables. Similarly, global temporary stored procedures are also scoped to the database level in Azure SQL Database. Local temporary tables (initiated with # table name) are also supported for Azure SQL Database and follow the same syntax and semantics that SQL Server uses. See the above section on Temporary Tables.

Important

This feature is available for Azure SQL Database.

Troubleshoot global temporary tables for Azure SQL Database

For troubleshooting tempdb, see How to Monitor tempdb use.

Note

Only a server admin can access the troubleshooting DMVs in Azure SQL Database.

Permissions for temporary objects

Any user can create global temporary objects. Users can only access their own objects, unless they receive additional permissions.

Partitioned tables

Before creating a partitioned table by using CREATE TABLE, you must first create a partition function to specify how the table becomes partitioned. A partition function is created by using CREATE PARTITION FUNCTION. Second, you must create a partition scheme to specify the filegroups that will hold the partitions indicated by the partition function. A partition scheme is created by using CREATE PARTITION SCHEME. Placement of PRIMARY KEY or UNIQUE constraints to separate filegroups can't be specified for partitioned tables. For more information, see Partitioned Tables and Indexes.

PRIMARY KEY constraints

  • A table can contain only one PRIMARY KEY constraint.

  • The index generated by a PRIMARY KEY constraint can't cause the number of indexes on the table to exceed 999 nonclustered indexes and 1 clustered index.

  • If CLUSTERED or NONCLUSTERED isn't specified for a PRIMARY KEY constraint, CLUSTERED is used if there are no clustered indexes specified for UNIQUE constraints.

  • All columns defined within a PRIMARY KEY constraint must be defined as NOT NULL. If nullability isn't specified, all columns participating in a PRIMARY KEY constraint have their nullability set to NOT NULL.

    Note

    For memory-optimized tables, the nullable key column is allowed.

  • If a primary key is defined on a CLR user-defined type column, the implementation of the type must support binary ordering. For more information, see CLR User-Defined Types.

UNIQUE constraints

  • If CLUSTERED or NONCLUSTERED isn't specified for a UNIQUE constraint, NONCLUSTERED is used by default.
  • Each UNIQUE constraint generates an index. The number of UNIQUE constraints can't cause the number of indexes on the table to exceed 999 nonclustered indexes and 1 clustered index.
  • If a unique constraint is defined on a CLR user-defined type column, the implementation of the type must support binary or operator-based ordering. For more information, see CLR User-Defined Types.

FOREIGN KEY constraints

  • When a value other than NULL is entered into the column of a FOREIGN KEY constraint, the value must exist in the referenced column; otherwise, a foreign key violation error message is returned.

  • FOREIGN KEY constraints are applied to the preceding column, unless source columns are specified.

  • FOREIGN KEY constraints can reference only tables within the same database on the same server. Cross-database referential integrity must be implemented through triggers. For more information, see CREATE TRIGGER.

  • FOREIGN KEY constraints can reference another column in the same table. This is referred to as a self-reference.

  • The REFERENCES clause of a column-level FOREIGN KEY constraint can list only one reference column. This column must have the same data type as the column on which the constraint is defined.

  • The REFERENCES clause of a table-level FOREIGN KEY constraint must have the same number of reference columns as the number of columns in the constraint column list. The data type of each reference column must also be the same as the corresponding column in the column list. The reference columns must be specified in the same order that was used when specifying the columns of the primary key or unique constraint on the referenced table.

  • CASCADE, SET NULL or SET DEFAULT can't be specified if a column of type timestamp is part of either the foreign key or the referenced key.

  • CASCADE, SET NULL, SET DEFAULT and NO ACTION can be combined on tables that have referential relationships with each other. If the Database Engine encounters NO ACTION, it stops and rolls back related CASCADE, SET NULL and SET DEFAULT actions. When a DELETE statement causes a combination of CASCADE, SET NULL, SET DEFAULT and NO ACTION actions, all the CASCADE, SET NULL and SET DEFAULT actions are applied before the Database Engine checks for any NO ACTION.

  • The Database Engine doesn't have a predefined limit on either the number of FOREIGN KEY constraints a table can contain that reference other tables, or the number of FOREIGN KEY constraints that are owned by other tables that reference a specific table.

    Nevertheless, the actual number of FOREIGN KEY constraints that can be used is limited by the hardware configuration and by the design of the database and application. We recommend that a table contain no more than 253 FOREIGN KEY constraints, and that it be referenced by no more than 253 FOREIGN KEY constraints. The effective limit for you may be more or less depending on the application and hardware. Consider the cost of enforcing FOREIGN KEY constraints when you design your database and applications.

  • FOREIGN KEY constraints aren't enforced on temporary tables.

  • FOREIGN KEY constraints can reference only columns in PRIMARY KEY or UNIQUE constraints in the referenced table or in a UNIQUE INDEX on the referenced table.

  • If a foreign key is defined on a CLR user-defined type column, the implementation of the type must support binary ordering. For more information, see CLR User-Defined Types.

  • Columns participating in a foreign key relationship must be defined with the same length and scale.

DEFAULT definitions

  • A column can have only one DEFAULT definition.

  • A DEFAULT definition can contain constant values, functions, SQL standard niladic functions, or NULL. The following table shows the niladic functions and the values they return for the default during an INSERT statement.

    SQL-92 niladic function Value returned
    CURRENT_TIMESTAMP Current date and time.
    CURRENT_USER Name of user performing an insert.
    SESSION_USER Name of user performing an insert.
    SYSTEM_USER Name of user performing an insert.
    USER Name of user performing an insert.
  • constant_expression in a DEFAULT definition can't refer to another column in the table, or to other tables, views, or stored procedures.

  • DEFAULT definitions can't be created on columns with a timestamp data type or columns with an IDENTITY property.

  • DEFAULT definitions can't be created for columns with alias data types if the alias data type is bound to a default object.

CHECK constraints

  • A column can have any number of CHECK constraints, and the condition can include multiple logical expressions combined with AND and OR. Multiple CHECK constraints for a column are validated in the order they are created.

  • The search condition must evaluate to a Boolean expression and can't reference another table.

  • A column-level CHECK constraint can reference only the constrained column, and a table-level CHECK constraint can reference only columns in the same table.

    CHECK CONSTRAINTS and rules serve the same function of validating the data during INSERT and UPDATE statements.

  • When a rule and one or more CHECK constraints exist for a column or columns, all restrictions are evaluated.

  • CHECK constraints can't be defined on text, ntext, or image columns.

Further constraint information

  • An index created for a constraint can't be dropped by using DROP INDEX; the constraint must be dropped by using ALTER TABLE. An index created for and used by a constraint can be rebuilt by using ALTER INDEX ... REBUILD. For more information, see Reorganize and Rebuild Indexes.
  • Constraint names must follow the rules for identifiers, except that the name can't start with a number sign (#). If constraint_name isn't supplied, a system-generated name is assigned to the constraint. The constraint name appears in any error message about constraint violations.
  • When a constraint is violated in an INSERT, UPDATE, or DELETE statement, the statement is ended. However, when SET XACT_ABORT is set to OFF, the transaction, if the statement is part of an explicit transaction, continues to be processed. When SET XACT_ABORT is set to ON, the whole transaction is rolled back. You can also use the ROLLBACK TRANSACTION statement with the transaction definition by checking the @@ERROR system function.
  • When ALLOW_ROW_LOCKS = ON and ALLOW_PAGE_LOCK = ON, row-, page-, and table-level locks are allowed when you access the index. The Database Engine chooses the appropriate lock and can escalate the lock from a row or page lock to a table lock. When ALLOW_ROW_LOCKS = OFF and ALLOW_PAGE_LOCK = OFF, only a table-level lock is allowed when you access the index.
  • If a table has FOREIGN KEY or CHECK CONSTRAINTS and triggers, the constraint conditions are evaluated before the trigger is executed.

For a report on a table and its columns, use sp_help or sp_helpconstraint. To rename a table, use sp_rename. For a report on the views and stored procedures that depend on a table, use sys.dm_sql_referenced_entities and sys.dm_sql_referencing_entities.

Nullability rules within a table definition

The nullability of a column determines whether that column can allow a null value (NULL) as the data in that column. NULL isn't zero or blank: NULL means no entry was made or an explicit NULL was supplied, and it typically implies that the value is either unknown or not applicable.

When you use CREATE TABLE or ALTER TABLE to create or alter a table, database and session settings influence and possibly override the nullability of the data type that is used in a column definition. We recommend that you always explicitly define a column as NULL or NOT NULL for noncomputed columns or, if you use a user-defined data type, that you allow the column to use the default nullability of the data type. Sparse columns must always allow NULL.

When column nullability isn't explicitly specified, column nullability follows the rules shown in the following table.

Column data type Rule
Alias data type The Database Engine uses the nullability that is specified when the data type was created. To determine the default nullability of the data type, use sp_help.
CLR user-defined type Nullability is determined according to the column definition.
System-supplied data type If the system-supplied data type has only one option, it takes precedence. timestamp data types must be NOT NULL. When any session settings are set ON by using SET:
ANSI_NULL_DFLT_ON = ON, NULL is assigned.
ANSI_NULL_DFLT_OFF = ON, NOT NULL is assigned.

When any database settings are configured by using ALTER DATABASE:
ANSI_NULL_DEFAULT_ON = ON, NULL is assigned.
ANSI_NULL_DEFAULT_OFF = ON, NOT NULL is assigned.

To view the database setting for ANSI_NULL_DEFAULT, use the sys.databases catalog view

When neither of the ANSI_NULL_DFLT options is set for the session and the database is set to the default (ANSI_NULL_DEFAULT is OFF), the default of NOT NULL is assigned.

If the column is a computed column, its nullability is always automatically determined by the Database Engine. To find out the nullability of this type of column, use the COLUMNPROPERTY function with the AllowsNull property.

Note

The SQL Server ODBC driver and SQL Server OLE DB driver both default to having ANSI_NULL_DFLT_ON set to ON. ODBC and OLE DB users can configure this in ODBC data sources, or with connection attributes or properties set by the application.

Data compression

System tables can't be enabled for compression. When you are creating a table, data compression is set to NONE, unless specified otherwise. If you specify a list of partitions or a partition that is out of range, an error will be generated. For a more information about data compression, see Data Compression.

To evaluate how changing the compression state will affect a table, an index, or a partition, use the sp_estimate_data_compression_savings stored procedure.

Permissions

Requires CREATE TABLE permission in the database and ALTER permission on the schema in which the table is being created.

If any columns in the CREATE TABLE statement are defined to be of a user-defined type, REFERENCES permission on the user-defined type is required.

If any columns in the CREATE TABLE statement are defined to be of a CLR user-defined type, either ownership of the type or REFERENCES permission on it is required.

If any columns in the CREATE TABLE statement have an XML schema collection associated with them, either ownership of the XML schema collection or REFERENCES permission on it is required.

Any user can create temporary tables in tempdb.

If the statement creates a ledger table, ENABLE LEDGER permission is required.

Examples

A. Create a PRIMARY KEY constraint on a column

The following example shows the column definition for a PRIMARY KEY constraint with a clustered index on the EmployeeID column of the Employee table. Because a constraint name isn't specified, the system supplies the constraint name.

CREATE TABLE dbo.Employee (
    EmployeeID INT PRIMARY KEY CLUSTERED
);

B. Use FOREIGN KEY constraints

A FOREIGN KEY constraint is used to reference another table. Foreign keys can be single-column keys or multicolumn keys. This following example shows a single-column FOREIGN KEY constraint on the SalesOrderHeader table that references the SalesPerson table. Only the REFERENCES clause is required for a single-column FOREIGN KEY constraint.

SalesPersonID INT NULL REFERENCES SalesPerson(SalesPersonID)

You can also explicitly use the FOREIGN KEY clause and restate the column attribute. The column name doesn't have to be the same in both tables.

FOREIGN KEY (SalesPersonID) REFERENCES SalesPerson(SalesPersonID)

Multicolumn key constraints are created as table constraints. In the AdventureWorks2022 database, the SpecialOfferProduct table includes a multicolumn PRIMARY KEY. The following example shows how to reference this key from another table; an explicit constraint name is optional.

CONSTRAINT FK_SpecialOfferProduct_SalesOrderDetail
    FOREIGN KEY (ProductID, SpecialOfferID)
    REFERENCES SpecialOfferProduct (ProductID, SpecialOfferID)

C. Use UNIQUE constraints

UNIQUE constraints are used to enforce uniqueness on nonprimary key columns. The following example enforces a restriction that the Name column of the Product table must be unique.

Name NVARCHAR(100) NOT NULL
UNIQUE NONCLUSTERED

D. Use DEFAULT definitions

Defaults supply a value (with the INSERT and UPDATE statements) when no value is supplied. For example, the AdventureWorks2022 database could include a lookup table listing the different jobs employees can fill in the company. Under a column that describes each job, a character string default could supply a description when an actual description isn't entered explicitly.

DEFAULT 'New Position - title not formalized yet'

In addition to constants, DEFAULT definitions can include functions. Use the following example to get the current date for an entry.

DEFAULT (GETDATE())

A niladic-function scan can also improve data integrity. To keep track of the user that inserted a row, use the niladic-function for USER. Don't enclose the niladic-functions with parentheses.

DEFAULT USER

E. Use CHECK constraints

The following example shows a restriction made to values that are entered into the CreditRating column of the Vendor table. The constraint is unnamed.

CHECK (CreditRating >= 1 and CreditRating <= 5)

This example shows a named constraint with a pattern restriction on the character data entered into a column of a table.

CONSTRAINT CK_emp_id CHECK (
    emp_id LIKE '[A-Z][A-Z][A-Z][1-9][0-9][0-9][0-9][0-9][FM]'
    OR emp_id LIKE '[A-Z]-[A-Z][1-9][0-9][0-9][0-9][0-9][FM]'
)

This example specifies that the values must be within a specific list or follow a specified pattern.

CHECK (
    emp_id IN ('1389', '0736', '0877', '1622', '1756')
    OR emp_id LIKE '99[0-9][0-9]'
)

F. Show the complete table definition

The following example shows the complete table definitions with all constraint definitions for table PurchaseOrderDetail created in the AdventureWorks2022 database. To run the sample, the table schema is changed to dbo.

CREATE TABLE dbo.PurchaseOrderDetail
(
    PurchaseOrderID int NOT NULL
        REFERENCES Purchasing.PurchaseOrderHeader(PurchaseOrderID),
    LineNumber smallint NOT NULL,
    ProductID int NULL
        REFERENCES Production.Product(ProductID),
    UnitPrice money NULL,
    OrderQty smallint NULL,
    ReceivedQty float NULL,
    RejectedQty float NULL,
    DueDate datetime NULL,
    rowguid uniqueidentifier ROWGUIDCOL NOT NULL
        CONSTRAINT DF_PurchaseOrderDetail_rowguid DEFAULT (NEWID()),
    ModifiedDate datetime NOT NULL
        CONSTRAINT DF_PurchaseOrderDetail_ModifiedDate DEFAULT (GETDATE()),
    LineTotal AS ((UnitPrice*OrderQty)),
    StockedQty AS ((ReceivedQty-RejectedQty)),
    CONSTRAINT PK_PurchaseOrderDetail_PurchaseOrderID_LineNumber
               PRIMARY KEY CLUSTERED (PurchaseOrderID, LineNumber)
               WITH (IGNORE_DUP_KEY = OFF)
)
ON [PRIMARY];

G. Create a table with an xml column typed to an XML schema collection

The following example creates a table with an xml column that is typed to XML schema collection HRResumeSchemaCollection. The DOCUMENT keyword specifies that each instance of the xml data type in column_name can contain only one top-level element.

CREATE TABLE HumanResources.EmployeeResumes
(
    LName nvarchar(25),
    FName nvarchar(25),
    Resume xml(DOCUMENT HumanResources.HRResumeSchemaCollection)
);

H. Create a partitioned table

The following example creates a partition function to partition a table or index into four partitions. Then, the example creates a partition scheme that specifies the filegroups in which to hold each of the four partitions. Finally, the example creates a table that uses the partition scheme. This example assumes the filegroups already exist in the database.

CREATE PARTITION FUNCTION myRangePF1 (int)
    AS RANGE LEFT FOR VALUES (1, 100, 1000);
GO

CREATE PARTITION SCHEME myRangePS1
    AS PARTITION myRangePF1
    TO (test1fg, test2fg, test3fg, test4fg);
GO

CREATE TABLE PartitionTable (col1 int, col2 char(10))
    ON myRangePS1 (col1);
GO

Based on the values of column col1 of PartitionTable, the partitions are assigned in the following ways.

Filegroup test1fg test2fg test3fg test4fg
Partition 1 2 3 4
Values col 1 <= 1 col1 > 1 AND col1 <= 100 col1 > 100 AND col1 <= 1,000 col1 > 1000

I. Use the UNIQUEIDENTIFIER data type in a column

The following example creates a table with a uniqueidentifier column. The example uses a PRIMARY KEY constraint to protect the table against users inserting duplicated values, and it uses the NEWSEQUENTIALID() function in the DEFAULT constraint to provide values for new rows. The ROWGUIDCOL property is applied to the uniqueidentifier column so that it can be referenced using the $ROWGUID keyword.

CREATE TABLE dbo.Globally_Unique_Data
(
    GUID UNIQUEIDENTIFIER
        CONSTRAINT Guid_Default DEFAULT
        NEWSEQUENTIALID() ROWGUIDCOL,
    Employee_Name VARCHAR(60)
    CONSTRAINT Guid_PK PRIMARY KEY (GUID)
);

J. Use an expression for a computed column

The following example shows the use of an expression ((low + high)/2) for calculating the myavg computed column.

CREATE TABLE dbo.mytable
(
    low INT,
    high INT,
    myavg AS (low + high)/2
);

K. Create a computed column based on a user-defined type column

The following example creates a table with one column defined as user-defined type utf8string, assuming that the type's assembly, and the type itself, have already been created in the current database. A second column is defined based on utf8string, and uses method ToString() of type(class) utf8string to compute a value for the column.

CREATE TABLE UDTypeTable
(
    u UTF8STRING,
    ustr AS u.ToString() PERSISTED
);

L. Use the USER_NAME function for a computed column

The following example uses the USER_NAME() function in the myuser_name column.

CREATE TABLE dbo.mylogintable
(
    date_in DATETIME,
    user_id INT,
    myuser_name AS USER_NAME()
);

M. Create a table that has a FILESTREAM column

The following example creates a table that has a FILESTREAM column Photo. If a table has one or more FILESTREAM columns, the table must have one ROWGUIDCOL column.

CREATE TABLE dbo.EmployeePhoto
(
    EmployeeId INT NOT NULL PRIMARY KEY,
    Photo VARBINARY(MAX) FILESTREAM NULL,
    MyRowGuidColumn UNIQUEIDENTIFIER NOT NULL ROWGUIDCOL UNIQUE DEFAULT NEWID()
);

N. Create a table that uses row compression

The following example creates a table that uses row compression.

CREATE TABLE dbo.T1
(
    c1 INT,
    c2 NVARCHAR(200)
)
WITH (DATA_COMPRESSION = ROW);

For additional data compression examples, see Data Compression.

O. Create a table that uses XML compression

Applies to: SQL Server 2022 (16.x) and later versions, Azure SQL Database, and Azure SQL Managed Instance.

The following example creates a table that uses XML compression.

CREATE TABLE dbo.T1
(
    c1 INT,
    c2 XML
)
WITH (XML_COMPRESSION = ON);

P. Create a table that has sparse columns and a column set

The following examples show to how to create a table that has a sparse column, and a table that has two sparse columns and a column set. The examples use the basic syntax. For more complex examples, see Use Sparse Columns and Use Column Sets.

This example creates a table that has a sparse column.

CREATE TABLE dbo.T1
(
    c1 INT PRIMARY KEY,
    c2 VARCHAR(50) SPARSE NULL
);

This example creates a table that has two sparse columns and a column set named CSet.

CREATE TABLE T1
(
    c1 INT PRIMARY KEY,
    c2 VARCHAR(50) SPARSE NULL,
    c3 INT SPARSE NULL,
    CSet XML COLUMN_SET FOR ALL_SPARSE_COLUMNS
);

Q. Create a system-versioned disk-based temporal table

Applies to: SQL Server 2016 (13.x) and later, and Azure SQL Database.

The following examples show how to create a temporal table linked to a new history table, and how to create a temporal table linked to an existing history table. The temporal table must have a primary key defined to be enabled for the table to be enabled for system versioning. For examples showing how to add or remove system versioning on an existing table, see System Versioning in Examples. For use cases, see Temporal Tables.

This example creates a new temporal table linked to a new history table.

CREATE TABLE Department
(
    DepartmentNumber CHAR(10) NOT NULL PRIMARY KEY CLUSTERED,
    DepartmentName VARCHAR(50) NOT NULL,
    ManagerID INT NULL,
    ParentDepartmentNumber CHAR(10) NULL,
    ValidFrom DATETIME2 GENERATED ALWAYS AS ROW START HIDDEN NOT NULL,
    ValidTo DATETIME2 GENERATED ALWAYS AS ROW END HIDDEN NOT NULL,
    PERIOD FOR SYSTEM_TIME (ValidFrom, ValidTo)
)
WITH (SYSTEM_VERSIONING = ON);

This example creates a new temporal table linked to an existing history table.

-- Existing table
CREATE TABLE Department_History
(
    DepartmentNumber CHAR(10) NOT NULL,
    DepartmentName VARCHAR(50) NOT NULL,
    ManagerID INT NULL,
    ParentDepartmentNumber CHAR(10) NULL,
    ValidFrom DATETIME2 NOT NULL,
    ValidTo DATETIME2 NOT NULL
);

-- Temporal table
CREATE TABLE Department
(
    DepartmentNumber CHAR(10) NOT NULL PRIMARY KEY CLUSTERED,
    DepartmentName VARCHAR(50) NOT NULL,
    ManagerID INT NULL,
    ParentDepartmentNumber CHAR(10) NULL,
    ValidFrom DATETIME2 GENERATED ALWAYS AS ROW START HIDDEN NOT NULL,
    ValidTo DATETIME2 GENERATED ALWAYS AS ROW END HIDDEN NOT NULL,
    PERIOD FOR SYSTEM_TIME (ValidFrom, ValidTo)
)
WITH (SYSTEM_VERSIONING = ON (HISTORY_TABLE = dbo.Department_History, DATA_CONSISTENCY_CHECK = ON));

R. Create a system-versioned memory-optimized temporal table

Applies to: SQL Server 2016 (13.x) and later, and Azure SQL Database.

The following example shows how to create a system-versioned memory-optimized temporal table linked to a new disk-based history table.

This example creates a new temporal table linked to a new history table.

CREATE SCHEMA History;
GO

CREATE TABLE dbo.Department
(
    DepartmentNumber CHAR(10) NOT NULL PRIMARY KEY NONCLUSTERED,
    DepartmentName VARCHAR(50) NOT NULL,
    ManagerID INT NULL,
    ParentDepartmentNumber CHAR(10) NULL,
    ValidFrom DATETIME2 GENERATED ALWAYS AS ROW START HIDDEN NOT NULL,
    ValidTo DATETIME2 GENERATED ALWAYS AS ROW END HIDDEN NOT NULL,
    PERIOD FOR SYSTEM_TIME (ValidFrom, ValidTo)
)
WITH
(
    MEMORY_OPTIMIZED = ON,
    DURABILITY = SCHEMA_AND_DATA,
    SYSTEM_VERSIONING = ON (HISTORY_TABLE = History.DepartmentHistory)
);

This example creates a new temporal table linked to an existing history table.

-- Existing table
CREATE TABLE Department_History
(
    DepartmentNumber CHAR(10) NOT NULL,
    DepartmentName VARCHAR(50) NOT NULL,
    ManagerID INT NULL,
    ParentDepartmentNumber CHAR(10) NULL,
    ValidFrom DATETIME2 NOT NULL,
    ValidTo DATETIME2 NOT NULL
);

-- Temporal table
CREATE TABLE Department
(
    DepartmentNumber CHAR(10) NOT NULL PRIMARY KEY CLUSTERED,
    DepartmentName VARCHAR(50) NOT NULL,
    ManagerID INT NULL,
    ParentDepartmentNumber CHAR(10) NULL,
    ValidFrom DATETIME2 GENERATED ALWAYS AS ROW START HIDDEN NOT NULL,
    ValidTo DATETIME2 GENERATED ALWAYS AS ROW END HIDDEN NOT NULL,
    PERIOD FOR SYSTEM_TIME (ValidFrom, ValidTo)
)
WITH
(
    SYSTEM_VERSIONING = ON (HISTORY_TABLE = dbo.Department_History, DATA_CONSISTENCY_CHECK = ON)
);

S. Create a table with encrypted columns

The following example creates a table with two encrypted columns. For more information, see Always Encrypted.

CREATE TABLE Customers (
    CustName NVARCHAR(60)
        ENCRYPTED WITH (
            COLUMN_ENCRYPTION_KEY = MyCEK,
            ENCRYPTION_TYPE = RANDOMIZED,
            ALGORITHM = 'AEAD_AES_256_CBC_HMAC_SHA_256'
        ),
    SSN VARCHAR(11) COLLATE Latin1_General_BIN2
        ENCRYPTED WITH (
            COLUMN_ENCRYPTION_KEY = MyCEK,
            ENCRYPTION_TYPE = DETERMINISTIC ,
            ALGORITHM = 'AEAD_AES_256_CBC_HMAC_SHA_256'
        ),
    Age INT NULL
);

T. Create an inline filtered index

Creates a table with an inline filtered index.

CREATE TABLE t1
(
    c1 INT,
    index IX1 (c1) WHERE c1 > 0
);

U. Create an inline index

The following shows how to use NONCLUSTERED inline for disk-based tables:

CREATE TABLE t1
(
    c1 INT,
    INDEX ix_1 NONCLUSTERED (c1)
);

CREATE TABLE t2
(
    c1 INT,
    c2 INT INDEX ix_1 NONCLUSTERED
);

CREATE TABLE t3
(
    c1 INT,
    c2 INT,
    INDEX ix_1 NONCLUSTERED (c1,c2)
);

V. Create a temporary table with an anonymously named compound primary key

Creates a table with an anonymously named compound primary key. This is useful to avoid run-time conflicts where two session-scoped temp tables, each in a separate session, use the same name for a constraint.

CREATE TABLE #tmp
(
    c1 INT,
    c2 INT,
    PRIMARY KEY CLUSTERED ([c1], [c2])
);
GO

If you explicitly name the constraint, the second session will generate an error such as:

Msg 2714, Level 16, State 5, Line 1
There is already an object named 'PK_#tmp' in the database.
Msg 1750, Level 16, State 1, Line 1
Could not create constraint or index. See previous errors.

The problem arises from the fact that while the temp table name is unique, the constraint names aren't.

W. Use global temporary tables in Azure SQL Database

Session A creates a global temp table ##test in Azure SQL Database testdb1 and adds one row

CREATE TABLE ##test (
    a INT,
    b INT
);

INSERT INTO ##test
VALUES (1, 1);

-- Obtain object ID for temp table ##test
SELECT OBJECT_ID('tempdb.dbo.##test') AS 'Object ID';

Here's the result set.

1253579504

Obtain global temp table name for a given object ID 1253579504 in tempdb (2)

SELECT name FROM tempdb.sys.objects WHERE object_id = 1253579504;

Here's the result set.

##test

Session B connects to Azure SQL Database testdb1 and can access table ##test created by session A

SELECT * FROM ##test;

Here's the result set.

1, 1

Session C connects to another database in Azure SQL Database testdb2 and wants to access ##test created in testdb1. This select fails due to the database scope for the global temp tables

SELECT * FROM ##test

Which generates the following error:

Msg 208, Level 16, State 0, Line 1
Invalid object name '##test'

Addressing system object in Azure SQL Database tempdb from current user database testdb1

SELECT * FROM tempdb.sys.objects;
SELECT * FROM tempdb.sys.columns;
SELECT * FROM tempdb.sys.database_files;

X. Enable Data Retention Policy on a table

The following example creates a table with data retention enabled and a retention period of one week. This example applies to Azure SQL Edge only.

CREATE TABLE [dbo].[data_retention_table]
(
  [dbdatetime2] datetime2(7),
  [product_code] int,
  [value] char(10)
)
WITH (DATA_DELETION = ON ( FILTER_COLUMN = [dbdatetime2], RETENTION_PERIOD = 1 WEEKS ))

Y. Create an updatable ledger table

The following example creates an updatable ledger table that isn't a temporal table with an anonymous history table (the system will generate the name of the history table) and the generated ledger view name. As the names of the required generated always columns and the additional columns in the ledger view aren't specified, the columns will have the default names.

CREATE SCHEMA [HR];
GO
CREATE TABLE [HR].[Employees]
(
    EmployeeID INT NOT NULL,
    Salary Money NOT NULL
)
WITH (SYSTEM_VERSIONING = ON, LEDGER = ON);
GO

The following example creates a table that is both a temporal table and an updatable ledger table, with an anonymous history table (with a name generated by the system), the generated ledger view name and the default names of the generated always columns and the additional ledger view columns.

CREATE SCHEMA [HR];
GO
CREATE TABLE [HR].[Employees]
(
    EmployeeID INT NOT NULL PRIMARY KEY,
    Salary Money NOT NULL,
    ValidFrom DATETIME2 GENERATED ALWAYS AS ROW START HIDDEN NOT NULL,
    ValidTo DATETIME2 GENERATED ALWAYS AS ROW END HIDDEN NOT NULL,
    PERIOD FOR SYSTEM_TIME (ValidFrom, ValidTo)
)
WITH (SYSTEM_VERSIONING = ON, LEDGER = ON);
GO

The following example creates a table that is both a temporal table and an updatable ledger table with the explicitly named history table, the user-specified name of the ledger view, and the user-specified names of generated always columns and additional columns in the ledger view.

CREATE SCHEMA [HR];
GO
CREATE TABLE [HR].[Employees]
(
    EmployeeID INT NOT NULL PRIMARY KEY,
    Salary Money NOT NULL,
    StartTransactionId BIGINT GENERATED ALWAYS AS TRANSACTION_ID START HIDDEN NOT NULL,
    EndTransactionId BIGINT GENERATED ALWAYS AS TRANSACTION_ID END HIDDEN NULL,
    StartSequenceNumber BIGINT GENERATED ALWAYS AS SEQUENCE_NUMBER START HIDDEN NOT NULL,
    EndSequenceNumber BIGINT GENERATED ALWAYS AS SEQUENCE_NUMBER END HIDDEN NULL,
    ValidFrom DATETIME2 GENERATED ALWAYS AS ROW START HIDDEN NOT NULL,
    ValidTo DATETIME2 GENERATED ALWAYS AS ROW END HIDDEN NOT NULL,
    PERIOD FOR SYSTEM_TIME (ValidFrom, ValidTo)
)
WITH (
    SYSTEM_VERSIONING = ON (HISTORY_TABLE = [HR].[EmployeesHistory]),
    LEDGER = ON (
        LEDGER_VIEW = [HR].[EmployeesLedger] (
            TRANSACTION_ID_COLUMN_NAME = TransactionId,
            SEQUENCE_NUMBER_COLUMN_NAME = SequenceNumber,
            OPERATION_TYPE_COLUMN_NAME = OperationId,
            OPERATION_TYPE_DESC_COLUMN_NAME = OperationTypeDescription
        )
    )
);
GO

The following example creates an append-only ledger table with the generated names of the ledger view and the columns in the ledger view.

CREATE SCHEMA [AccessControl];
GO
CREATE TABLE [AccessControl].[KeyCardEvents]
(
    EmployeeID INT NOT NULL,
    AccessOperationDescription NVARCHAR (MAX) NOT NULL,
    [Timestamp] Datetime2 NOT NULL,
    StartTransactionId BIGINT GENERATED ALWAYS AS TRANSACTION_ID START HIDDEN NOT NULL,
    StartSequenceNumber BIGINT GENERATED ALWAYS AS SEQUENCE_NUMBER START HIDDEN NOT NULL
)
WITH (
    LEDGER = ON (
        LEDGER_VIEW = [AccessControl].[KeyCardEventsLedger] (
            TRANSACTION_ID_COLUMN_NAME = TransactionId,
            SEQUENCE_NUMBER_COLUMN_NAME = SequenceNumber,
            OPERATION_TYPE_COLUMN_NAME = OperationId,
            OPERATION_TYPE_DESC_COLUMN_NAME = OperationTypeDescription
        ),
        APPEND_ONLY = ON
    )
);
GO

The following example creates a ledger database in Azure SQL Database and an updatable ledger table using the default settings. Creating an updatable ledger table in a ledger database doesn't require using WITH (SYSTEM_VERSIONING = ON, LEDGER = ON);.

CREATE DATABASE MyLedgerDB ( EDITION = 'GeneralPurpose' ) WITH LEDGER = ON;
GO

CREATE SCHEMA [HR];
GO

CREATE TABLE [HR].[Employees]
(
    EmployeeID INT NOT NULL,
    Salary Money NOT NULL
)
GO