On this page

Show all

Help Center/ GaussDB/ Developer Guide(Distributed_V2.0-2.x)/ SQL Reference/ SQL Syntax/ ALTER TABLE

ALTER TABLE

Updated on 2025-02-27 GMT+08:00

View PDF

Description

Alters tables, including modifying table definitions, renaming tables, renaming specified columns in tables, renaming table constraints, setting table schemas, enabling or disabling row-level security policies, and adding or updating multiple columns.

Precautions

The owner of a table, users granted with the ALTER permission on the table, or users granted with the ALTER ANY TABLE permission can run the ALTER TABLE command. System administrators have the permission to run the command by default. To modify the owner or schema of a table, you must be the table owner or a system administrator and a member of the new owner role.
The tablespace of a partitioned table cannot be modified, but the tablespace of a partition can be modified.
The storage parameter ORIENTATION cannot be modified.
Currently, SET SCHEMA can only set schemas to user schemas. It cannot set a schema to a system internal schema.
The distribution key (or column) of a table cannot be modified.

Auto-increment columns cannot be added, or a column whose DEFAULT value contains the nextval() expression cannot be added.
Row-level security cannot be enabled for foreign tables and temporary tables.
When you delete a PRIMARY KEY constraint by constraint name, the NOT NULL constraint is not deleted. If necessary, manually delete the NOT NULL constraint.
When JDBC is used, the DEFAULT value can be set through PrepareStatement.
If you add a column using ADD COLUMN, all existing rows in the table are initialized to the column's default value (NULL if no DEFAULT value is specified).
If no DEFAULT value is specified for the new column, NULL is used, and no full table update is triggered.

If the new column has a DEFAULT value, the column must meet all the following requirements. Otherwise, the entire table is updated, leading to additional overheads and affecting online services.

1. The data type is BOOL, BYTEA, SMALLINT, BIGINT, SMALLINT, INTEGER, NUMERIC, FLOAT, DOUBLE PRECISION, CHAR, VARCHAR, TEXT, TIMESTAMPTZ, TIMESTAMP, DATE, TIME, TIMETZ, or INTERVAL.

2. The length of the DEFAULT value of the added column cannot exceed 128 bytes.

3. The DEFAULT value of the added column does not contain the volatile function.

4. The DEFAULT value is required and cannot be NULL.

If you are not sure whether condition 3 is met, check whether the provolatile attribute of the function in the PG_RPOC system catalog is 'v'.

Syntax

Modify the definition of a table.

      
         ALTER TABLE [ IF EXISTS ] { table_name [*] | ONLY table_name | ONLY ( table_name ) }
    action [, ... ];

The table operation ACTION can be one of the following clauses:

       
        
          
          column_clause
    | ADD table_constraint [ NOT VALID ]
    | ADD table_constraint_using_index
    | VALIDATE CONSTRAINT constraint_name
    | DROP CONSTRAINT [ IF EXISTS ]  constraint_name [ RESTRICT | CASCADE ]
    | CLUSTER ON index_name
    | SET WITHOUT CLUSTER
    | SET ( {storage_parameter = value} [, ... ] )
    | RESET ( storage_parameter [, ... ] )
    | OWNER TO new_owner
    | SET TABLESPACE new_tablespace
    | TO { GROUP groupname | NODE ( nodename [, ... ] ) }
    | ADD NODE ( nodename [, ... ] )
    | DELETE NODE ( nodename [, ... ] )
    | UPDATE SLICE LIKE table_name
    | DISABLE TRIGGER [ trigger_name | ALL | USER ]
    | ENABLE TRIGGER [ trigger_name | ALL | USER ]
    | ENABLE REPLICA TRIGGER trigger_name
    | ENABLE ALWAYS TRIGGER trigger_name
    | DISABLE/ENABLE [ REPLICA | ALWAYS ] RULE
    | DISABLE ROW LEVEL SECURITY
    | ENABLE ROW LEVEL SECURITY
    | FORCE ROW LEVEL SECURITY
    | NO FORCE ROW LEVEL SECURITY
    
    | INHERIT parent_table
    | NO INHERIT parent_table
    | DISTRIBUTE BY { REPLICATION | { [ HASH ] ( column_name ) } }
    | OF type_name
    | NOT OF
    | REPLICA IDENTITY { DEFAULT | USING INDEX index_name | FULL | NOTHING }

         

       
      

NOTE:

ADD table_constraint [ NOT VALID ]
Adds a table constraint.
ADD table_constraint_using_index
Adds a primary key constraint or unique constraint to a table based on the existing unique index.
VALIDATE CONSTRAINT constraint_name
Validates a check-class constraint created with the NOT VALID option, and scans the entire table to ensure that all rows meet the constraint. Nothing happens if the constraint is already marked valid.
DROP CONSTRAINT [ IF EXISTS ] constraint_name [ RESTRICT | CASCADE ]
Drops a table constraint.
CLUSTER ON index_name
Selects the default index for future CLUSTER operations. Actually, the table is not re-clustered.
SET WITHOUT CLUSTER
Deletes the most recently used CLUSTER index from the table. This affects future CLUSTER operations that do not specify an index.
SET ( {storage_parameter = value} [, ... ] )
Changes one or more storage parameters for the table.
RESET ( storage_parameter [, ... ] )
Resets one or more storage parameters to their defaults. As with SET, a table rewrite might be needed to update the table entirely.
OWNER TO new_owner
Changes the owner of a table, sequence, or view to the specified user.
SET TABLESPACE new_tablespace
Changes the table's tablespace to the specified tablespace and moves the data files associated with the table to the new tablespace. Indexes on the table, if any, are not moved; but they can be moved separately with additional SET TABLESPACE option in ALTER INDEX.
TO { GROUP groupname | NODE ( nodename [, ... ] ) }
The syntax is only available in extended mode (when GUC parameter support_extended_features is on). Exercise caution when enabling the mode. It is mainly used for tools like internal dilatation tools. Common users should not use the mode. This command only modifies the logical mapping relationship of the table distribution nodes and does not migrate the table's metadata and data on the DN.
ADD NODE ( nodename [, ... ] )
It is only available for internal scale-out tools. Common users should not use the syntax.
DELETE NODE ( nodename [, ... ] )
It is only available for internal scale-in tools. Common users should not use the syntax.
UPDATE SLICE LIKE table_name
This syntax is used by internal scaling tools and cannot be used by common users.
DISABLE TRIGGER [ trigger_name | ALL | USER ]
Disables a single trigger specified by trigger_name, disables all triggers, or disables only user triggers (excluding internally generated constraint triggers, for example, deferrable unique constraint triggers and exclusion constraints triggers).

Exercise caution when using this function because data integrity cannot be ensured as expected if the triggers are not executed.
| ENABLE TRIGGER [ trigger_name | ALL | USER ]
Enables a single trigger specified by trigger_name, enables all triggers, or enables only user triggers.
| ENABLE REPLICA TRIGGER trigger_name
Determines that the trigger firing mechanism is affected by the configuration variable session_replication_role. When the replication role is origin (default value) or local, a simple trigger is fired.

When ENABLE REPLICA is configured for a trigger, it is fired only when the session is in replica mode.
| ENABLE ALWAYS TRIGGER trigger_name
Determines that all triggers are fired regardless of the current replication mode.
| { DISABLE | ENABLE } [ REPLICA | ALWAYS ] RULE
Enables or disables a rule for tables. Disabled rules are still visible in the system, but are not applied during query rewriting. The ON SELECT rule cannot be disabled because it is related to the view implementation. Rules configured as ENABLE REPLICA are enabled only when the session is in replica mode, while those configured as ENABLE ALWAYS can be enabled regardless of the replica mode. Rule triggering is also affected by configuration variable session_replication_role, which is similar to the preceding trigger setting.
| { DISABLE | ENABLE } ROW LEVEL SECURITY
Enables or disables row-level security for a table.

If row-level security is enabled for a data table but no row-level security policy is defined, the row-level access to the data table is not affected. If row-level security for a table is disabled, the row-level access to the table is not affected even if a row-level security policy has been defined. For details, see CREATE ROW LEVEL SECURITY POLICY.
| { NO FORCE | FORCE } ROW LEVEL SECURITY
Forcibly enables or disables row-level security for a table.

By default, the table owner is not affected by the row-level security feature. However, if row-level security is forcibly enabled, the table owner (excluding system administrators) will be affected. System administrators are not affected by any row-level security policies.

INHERIT parent_table
Adds the target data table to a specified parent data table as a new child data table. After that, the query for the parent data table will contain the data in the target data table. Before being added as a child data table, the target data table must contain all the columns in the parent data table. These columns must have matching data categories, and if they have NOT NULL constraints in the parent data table, they must also have NOT NULL constraints in the child data table. For all CHECK constraints in the parent data table, there must be corresponding constraints in the child data table, unless the parent data table is marked as non-inheritable.
NO INHERIT parent_table
Generates the target data table from the child data table of a specified parent data table. Queries for the parent data table will no longer contain records generated from the target data table.

DISTRIBUTE BY { REPLICATION | { [ HASH ] ( column_name ) } }
Specifies how the table is distributed or replicated between DNs.

OF type_name
Joins a table to a composite type, which is similar to table creation by using the CREATE TABLE OF option. The name and type of a table column must exactly match those defined in the composite type, but the OID system column can be different. The table cannot be inherited from any other table. These restrictions ensure that the CREATE TABLE OF option allows the same table definition.
NOT OF
Removes the association between a table and a type.

REPLICA IDENTITY { DEFAULT | USING INDEX index_name | FULL | NOTHING }
Specifies the record level of old tuples in UPDATE and DELETE statements on a table in logical replication scenarios.
- DEFAULT records the old value of the primary key column. If there is no primary key, DEFAULT does not record the old value.
- USING INDEX records the old values of columns covered by the named indexes. These values must be unique, non-local, and non-deferrable, and contain the values of columns marked NOT NULL.
- FULL records the old values of all columns in the row.
- NOTHING does not record information in old rows.
In logical replication scenarios, when the UPDATE and DELETE statements of a table are parsed, the parsed old tuples consist of the information recorded in this method. For tables with primary keys, this option can be set to DEFAULT or FULL. For a table without a primary key, set this parameter to FULL. Otherwise, the old tuple will be parsed as empty during decoding. You are advised not to set this parameter to NOTHING in common scenarios because old tuples are always parsed as empty.

There are several clauses of column_clause:

        
         
           
           ADD [ COLUMN ] column_name data_type [ COLLATE collation ] [ column_constraint [ ... ] ]    
| MODIFY column_name data_type    
| MODIFY column_name [ CONSTRAINT constraint_name ] NOT NULL [ ENABLE ]
| MODIFY column_name [ CONSTRAINT constraint_name ] NULL
| DROP [ COLUMN ] [ IF EXISTS ] column_name [ RESTRICT | CASCADE ]    
| ALTER [ COLUMN ] column_name [ SET DATA ] TYPE data_type [ COLLATE collation ] [ USING expression ]    
| ALTER [ COLUMN ] column_name { SET DEFAULT expression | DROP DEFAULT }    
| ALTER [ COLUMN ] column_name { SET | DROP } NOT NULL    
| ALTER [ COLUMN ] column_name SET STATISTICS [PERCENT] integer    
| ADD STATISTICS (( column_1_name, column_2_name [, ...] ))    
| DELETE STATISTICS (( column_1_name, column_2_name [, ...] ))    
| ALTER [ COLUMN ] column_name SET ( {attribute_option = value} [, ... ] )    
| ALTER [ COLUMN ] column_name RESET ( attribute_option [, ... ] )    
| ALTER [ COLUMN ] column_name SET STORAGE { PLAIN | EXTERNAL | EXTENDED | MAIN }

          

        
       

NOTE:

ADD [ COLUMN ] column_name data_type [ COLLATE collation ] [ column_constraint [ ... ] ]
Adds a column to a table. If you add a column using ADD COLUMN, all existing rows in the table are initialized to the column's default value (NULL if no DEFAULT value is specified).
ADD ( { column_name data_type} [, ...] )
Adds columns in the table.
MODIFY ( { column_name data_type | column_name [ CONSTRAINT constraint_name ] NOT NULL [ ENABLE ] | column_name [ CONSTRAINT constraint_name ] NULL } [, ...] )
Modifies the data type of an existing column in the table. Running this command will clear the statistics of this column. You are advised to collect the statistics of this column again after the modification.
DROP [ COLUMN ] [ IF EXISTS ] column_name [ RESTRICT | CASCADE ]
Drops a column from a table. Indexes and constraints related to the column are automatically dropped. If an object not belonging to the table depends on the column (for example, view), CASCADE must be specified.

The DROP COLUMN statement does not physically remove the column, but simply makes it invisible to SQL operations. Subsequent INSERT and UPDATE operations on the table will store a NULL value for the column. Therefore, dropping a column takes a short period of time but does not immediately release the tablespace on the disks, because the space occupied by the dropped column is not recycled. The space will be recycled when VACUUM is executed.
ALTER [ COLUMN ] column_name [ SET DATA ] TYPE data_type [ COLLATE collation ] [ USING expression ]
Alters the type of a column in a table. Indexes and simple table constraints on the column will automatically use the new data type by reparsing the originally supplied expression.

If the original data type of a column and the modified data type are binary compatible, you do not need to rewrite the entire table when running this statement. In other scenarios, the entire table is rewritten. You can check whether the original type and target type are binary compatible in the PG_CAST system catalog. If castmethod is 'b', they are binary compatible. For example, if the data type of the source table is text and is converted to int, table rewriting is triggered. If it is converted to clob, table rewriting is not triggered. If table rewriting is triggered, the deleted space on the table is recycled immediately.

Running this command will clear the statistics of this column. You are advised to collect the statistics of this column again after the modification.
ALTER [ COLUMN ] column_name { SET DEFAULT expression | DROP DEFAULT }
Sets or drops the default value for a column. The default values only apply to subsequent INSERT operations; they do not cause rows already in the table to change. Defaults can also be created for views, in which case they are inserted into INSERT statements on the view before the view's ON INSERT rule is applied.
ALTER [ COLUMN ] column_name { SET | DROP } NOT NULL
Alters whether a column is marked to allow NULL values or to reject NULL values. You can only use SET NOT NULL when the column contains non-NULL values.
ALTER [ COLUMN ] column_name SET STATISTICS [PERCENT] integer
Specifies the per-column statistics-collection target for subsequent ANALYZE operations. The target can be set in the range from 0 to 10000. Set it to –1 to revert to using the default system statistics target.
ALTER [ COLUMN ] column_name SET ( {attribute_option = value} [, ... ] )
ALTER [ COLUMN ] column_name RESET ( attribute_option [, ... ] )

Sets or resets per-attribute options.

Currently, the only defined per-attribute options are n_distinct and n_distinct_inherited. n_distinct affects statistics of a table, while n_distinct_inherited affects the statistics of the table and its subtables. Currently, only SET/RESET n_distinct is supported, and SET/RESET n_distinct_inherited is forbidden.
ALTER [ COLUMN ] column_name SET STORAGE { PLAIN | EXTERNAL | EXTENDED | MAIN }
Sets the storage mode for a column. This clause specifies whether this column is held inline or in a secondary TOAST table, and whether the data should be compressed. Only row-store tables can be set. SET STORAGE itself does not change anything in the table. It sets the strategy to be pursued during future table updates.

column_constraint is as follows:

          
             [ CONSTRAINT constraint_name ]
    { NOT NULL |
      NULL |
      CHECK ( expression ) |
      DEFAULT default_expr  |
      UNIQUE index_parameters |
      PRIMARY KEY index_parameters |
      REFERENCES reftable [ ( refcolumn ) ] [ MATCH FULL |  MATCH PARTIAL | MATCH SIMPLE ]
        [ ON DELETE action ] [ ON UPDATE action ] }
    [ DEFERRABLE | NOT DEFERRABLE ][ INITIALLY DEFERRED | INITIALLY IMMEDIATE ]

table_constraint_using_index used to add the primary key constraint or unique constraint based on the unique index is as follows:

        
           [ CONSTRAINT constraint_name ]
    { UNIQUE | PRIMARY KEY } USING INDEX index_name
    [ DEFERRABLE | NOT DEFERRABLE ][ INITIALLY DEFERRED | INITIALLY IMMEDIATE ]

table_constraint is as follows:

        
           [ CONSTRAINT constraint_name ]
    { CHECK ( expression ) |
      UNIQUE ( column_name [, ... ] ) index_parameters |
      PRIMARY KEY ( column_name [, ... ] ) index_parameters |
      PARTIAL CLUSTER KEY ( column_name [, ... ] )}
    [ DEFERRABLE | NOT DEFERRABLE ][ INITIALLY DEFERRED | INITIALLY IMMEDIATE ]

index_parameters is as follows:

         
            [ WITH ( {storage_parameter = value} [, ... ] ) ]
    [ USING INDEX TABLESPACE tablespace_name ]

Rename a table. The renaming does not affect stored data.

     
        ALTER TABLE [ IF EXISTS ] table_name 
    RENAME TO new_table_name;

Rename the specified column in the table.

     
        ALTER TABLE [ IF EXISTS ] { table_name [*] | ONLY table_name | ONLY ( table_name )}
    RENAME [ COLUMN ] column_name TO new_column_name;

Rename the constraint of the table.

     
        ALTER TABLE [ IF EXISTS ] { table_name [*] | ONLY table_name | ONLY ( table_name ) }
    RENAME CONSTRAINT constraint_name TO new_constraint_name;

Set the schema of the table.
1 2

ALTER TABLE [ IF EXISTS ] table_name SET SCHEMA new_schema;
NOTE:
- The schema setting moves the table into another schema. Associated indexes and constraints owned by table columns are migrated as well. Currently, the schema for sequences cannot be changed. If the table has sequences, delete the sequences, and create them again or delete the ownership between the table and sequences. In this way, the table schema can be changed.
- To change the schema of a table, you must also have the CREATE permission on the new schema. To add the table as a new child of a parent table, you must own the parent table as well. To alter the owner, you must also be a direct or indirect member of the new owning role, and that role must have the CREATE permission on the table's schema. These restrictions enforce that the user can only rebuild and delete the table. However, a system administrator can alter the ownership of any table anyway.
- All the actions except for RENAME and SET SCHEMA can be combined into a list of multiple alterations to apply in parallel. For example, it is possible to add several columns or alter the type of several columns in a single statement. This is useful with large tables, since only one pass over the tables need be made.
- Adding a CHECK or NOT NULL constraint will scan the table to validate that existing rows meet the constraint.
- Adding a column with a non-null default or changing the type of an existing column will rewrite the entire table. Rewriting a large table may take much time and temporarily needs doubled disk space.

Add columns.

     
        ALTER TABLE [ IF EXISTS ] table_name
    ADD ( { column_name data_type [ COLLATE collation ] [ column_constraint [ ... ] ]} [, ...] );

Update columns.

     
        ALTER TABLE [ IF EXISTS ] table_name 
    MODIFY ( { column_name data_type | column_name [ CONSTRAINT constraint_name ] NOT NULL [ ENABLE ] | column_name [ CONSTRAINT constraint_name ] NULL } [, ...] );

Parameters

IF EXISTS
Sends a notice instead of an error if no tables have identical names. The notice prompts that the table you are querying does not exist.
table_name [*] | ONLY table_name | ONLY ( table_name )
table_name is the name of the table that you need to modify.

If ONLY is specified, only the table is modified. If ONLY is not specified, the table and all subtables are modified. You can add the asterisk (*) option following the table name to specify that all subtables are scanned, which is the default operation.
constraint_name
Specifies the name of an existing constraint to drop.
index_name
Specifies the name of an index.
storage_parameter
Specifies the name of a storage parameter.

The following options are added for online scaling:
- append_mode (enumerated type)
  Scales out a table online or offline, or stops scaling it. You can modify certain content in the table during online scaling but cannot do so during offline scaling.
  
  To modify a table that is being scaled, append new data so that they can be recorded as incremental data.
  - on: scales out a table online. New data will be appended.
  - off: stops scaling. New data will be written in normal mode, and options for online scaling will not be displayed in pg_class.reloptions.
  - read_only: scales a table offline, during which no other operations can be performed on the table.
  - end_catchup: reports errors for the write service in the last round of data increment. The read service is executed normally.
- rel_cn_oid (OID type)
  Records the OID of tables on the current CN to generate delete_delta on the DNs.
  
  If append_mode is set to on, rel_cn_oid must be specified.
  
  The append_mode and rel_cn_oid parameters are used only by online scaling tools.
- exec_step (integer)
  Records resumable transmission steps in relOptions of the temporary table.
  
  Value range: [1,4]
  
  It can be used only for data redistribution.
- create_time (long integer)
  Records the time when the temporary table is created during resumable transmission in relOptions of the temporary table.
  
  It can be used only for data redistribution.
- wait_clean_cbi (string type)
  Specifies whether the current global index contains the residual tuple generated during bucket migration for scaling. After scaling, wait_clean_cbi is set to y. After the residual tuple is cleared in the vacuum process, wait_clean_cbi is set to n.
  
  This option is used only in scaling tools.
The following option is added for creating an index:
- parallel_workers (int type)
  Number of bgworker threads started when an index is created. For example, value 2 indicates that two bgworker threads are started to create indexes concurrently.
  
  Value range: [0,32]. The value 0 indicates that this function is disabled.
  
  Default value: If this parameter is not set, the concurrent index creation function is disabled.
The following option is added to the replication table:
- primarynode (Boolean type)
  Default value: off
  
  When primarynode is set to on, the primary node is selected for the replication table. Generally, the primary node is the first node recorded in the nodeoids column in the pgxc_class system catalog. When the IUD operation is performed on the replication table, the operation is delivered to the primary node first. After the result is received, the operation is delivered to other DNs.
new_owner
Specifies the name of the new table owner.
new_tablespace
Specifies the new name of the tablespace to which the table belongs.
column_name, column_1_name, column_2_name
Specifies the name of a new or existing column.
data_type
Specifies the type of a new column or a new type of an existing column.
collation
Specifies the collation rule name of a column. The optional COLLATE clause specifies a collation for the new column; if omitted, the collation is the default for the new column. You can run the select * from pg_collation command to query collation rules from the pg_collation system catalog. The default collation rule is the row starting with default in the query result.
USING expression
Specifies how to compute the new column value from the old; if omitted, the default conversion is an assignment cast from old data type to new. A USING clause must be provided if there is no implicit or assignment cast from the old to new type.

NOTE:

USING in ALTER TYPE can specify any expression involving the old values of the row; that is, it can refer to any columns other than the one being cast. This allows general casting to be done with the ALTER TYPE syntax. Because of this flexibility, the USING expression is not applied to the column's default value (if any); the result might not be a constant expression as required for a default. This means that when there is no implicit or assignment cast from old to new type, ALTER TYPE might fail to convert the default even though a USING clause is supplied. In such cases, drop the default with DROP DEFAULT, perform ALTER TYPE, and then use SET DEFAULT to add a suitable new default. Similar considerations apply to indexes and constraints involving the column.
NOT NULL | NULL
Sets whether the column allows null values.
ENABLE
Specifies that the constraint is enabled. By default, the constraint is enabled.
integer
Specifies the constant value of a signed integer. When using PERCENT, the range of integer is from 0 to 100.
attribute_option
Specifies an attribute option.
PLAIN | EXTERNAL | EXTENDED | MAIN
Specifies a column-store mode.
- PLAIN must be used for fixed-length values (such as integers). It must be inline and uncompressed.
- MAIN is for inline, compressible data.
- EXTERNAL is for external, uncompressed data. Use of EXTERNAL will make substring operations on text and bytea values run faster, at the penalty of increased storage space.
- EXTENDED is for external, compressed data. EXTENDED is the default for most data types that support non-PLAIN storage.
CHECK ( expression )
New rows or rows to be updated must satisfy for an expression to be true. If any row produces a false result, an error is raised and the database is not modified.

A check constraint specified as a column constraint should reference only the column's values, while an expression in a table constraint can reference multiple columns.

Currently, CHECK ( expression ) does not include subqueries and cannot use variables apart from the current column.
DEFAULT default_expr
Assigns a default data value to a column.

The data type of the default expression must match that of the column.

The default expression will be used in any INSERT operation that does not specify a value for the column. If there is no default value for a column, then the default value is NULL.
UNIQUE index_parameters
UNIQUE ( column_name [, ... ] ) index_parameters

Specifies that a group of one or more columns of a table can contain only unique values.
PRIMARY KEY index_parameters
PRIMARY KEY ( column_name [, ... ] ) index_parameters

Specifies that a column or columns of a table can contain only unique (non-duplicate) and non-NULL values.
DEFERRABLE | NOT DEFERRABLE | INITIALLY DEFERRED | INITIALLY IMMEDIATE
Sets whether the constraint can be deferrable.
- DEFERRABLE: deferrable to the end of the transaction and checked using SET CONSTRAINTS.
- NOT DEFERRABLE: checks immediately after the execution of each command.
- INITIALLY IMMEDIATE: checks immediately after the execution of each statement.
- INITIALLY DEFERRED: checks when the transaction ends.
WITH ( {storage_parameter = value} [, ... ] )
Specifies an optional storage parameter for a table or an index.
tablespace_name
Specifies the name of the tablespace where the index locates.
new_table_name
Specifies the new table name.
new_column_name
Specifies the new name of a specific column in a table.
new_constraint_name
Specifies the new name of a table constraint.
new_schema
Specifies the new schema name.
CASCADE
Automatically drops objects that depend on the dropped column or constraint (for example, views referencing the column).
RESTRICT
Refuses to drop the column or constraint if there are any dependent objects. This is the default behavior.
schema_name
Specifies the schema name of a table.

Examples of Altering a Table

Rename a table.

openGauss=# CREATE TABLE aa(c1 int, c2 int);
openGauss=# ALTER TABLE IF EXISTS aa RENAME TO test_alt1;

Alter the schema of a table.

-- Create test_schema.
openGauss=# CREATE SCHEMA test_schema;
-- Change the schema of the test_alt1 table to test_schema.
openGauss=# ALTER TABLE test_alt1 SET SCHEMA test_schema;
-- Query table information.
openGauss=# SELECT schemaname,tablename FROM pg_tables WHERE tablename = 'test_alt1';
 schemaname  | tablename 
-------------+-----------
 test_schema | test_alt1
(1 row)

Change the owner of a table.

-- Create test_user.
openGauss=# CREATE USER test_user PASSWORD 'XXXXXXXXXX';
-- Change the owner of the test_alt1 table to test_user.
openGauss=# ALTER TABLE IF EXISTS test_schema.test_alt1 OWNER TO test_user;
-- Query.
openGauss=# SELECT tablename, schemaname, tableowner FROM pg_tables WHERE tablename = 'test_alt1';
 tablename | schemaname  | tableowner 
-----------+-------------+------------
 test_alt1 | test_schema | test_user
(1 row)

Alter the tablespace of a table.

-- Create the tbs_data1 tablespace.
openGauss=# CREATE TABLESPACE tbs_data1 RELATIVE LOCATION 'tablespace1/tbs_data1';
-- Change the tablespace of the test_alt1 table to tbs_data1.
openGauss=# ALTER TABLE test_schema.test_alt1 SET TABLESPACE tbs_data1;
-- Query.
openGauss=# SELECT tablename, tablespace FROM pg_tables WHERE tablename = 'test_alt1';
 tablename | tablespace 
-----------+------------
 test_alt1 | tbs_data1
(1 row)

-- Drop.
openGauss=# DROP TABLE test_schema.test_alt1; 
openGauss=# DROP TABLESPACE tbs_data1; 
openGauss=# DROP SCHEMA test_schema; 
openGauss=# DROP USER test_user;

Examples of Changing a Column

Change column names.

-- Create a table.
openGauss=# CREATE TABLE test_alt2(c1 INT,c2 INT);
-- Change column names.
openGauss=# ALTER TABLE test_alt2 RENAME c1 TO id;
openGauss=# ALTER TABLE test_alt2 RENAME COLUMN c2 to areaid; 
-- Query.
openGauss=# \d test_alt1
   Table "public.test_alt1"
 Column |  Type   | Modifiers 
--------+---------+-----------
 id     | integer | 
 areaid | integer |

Add a column.

-- Add a column to the test_alt1 table.
openGauss=# ALTER TABLE IF EXISTS test_alt2 ADD COLUMN name VARCHAR(20);
-- Query.
openGauss=# \d test_alt2
           Table "public.test_alt1"
  Column  |         Type          | Modifiers 
----------+-----------------------+-----------
 id       | integer               | 
 areacode | integer               | 
 name     | character varying(20) |

Alter the data type of a column.

-- Change the type of the name column in the test_alt1 table.
openGauss=# ALTER TABLE test_alt1 MODIFY name VARCHAR(50);
-- Query.
openGauss=# \d test_alt1
          Table "public.test_alt2"
 Column |         Type          | Modifiers 
--------+-----------------------+-----------
 id     | integer               | 
 areaid | integer               | 
 name   | character varying(50) |

-- Change the type of the name column in the test_alt1 table.
openGauss=# ALTER TABLE test_alt2 ALTER COLUMN name TYPE VARCHAR(25);
-- Query.
openGauss=# \d test_alt2
          Table "public.test_alt2"
 Column |         Type          | Modifiers 
--------+-----------------------+-----------
 id     | integer               | 
 areaid | integer               | 
 name   | character varying(25) |

Drop a column.

-- Drop the areaid column from test_alt1.
openGauss=# ALTER TABLE test_alt2 DROP COLUMN areaid;
-- Query.
openGauss=# \d test_alt2
          Table "public.test_alt2"
 Column |         Type          | Modifiers 
--------+-----------------------+-----------
 id     | integer               | 
 name   | character varying(25) |

Modify the column-store mode.

-- View table details.
openGauss=# \d+ test_alt2
                              Table "public.test_alt2"
 Column |         Type          | Modifiers | Storage  | Stats target | Description 
--------+-----------------------+-----------+----------+--------------+-------------
 id     | integer               |           | plain    |              | 
 name   | character varying(25) |           | extended |              | 
Has OIDs: no
Options: orientation=row, storage_type=USTORE

-- Change the storage mode of the name column in the test_alt2 table.
openGauss=# ALTER TABLE test_alt2 ALTER COLUMN name SET STORAGE PLAIN;

-- Query.
openGauss=# \d+ test_alt2
                             Table "public.test_alt2"
 Column |         Type          | Modifiers | Storage | Stats target | Description 
--------+-----------------------+-----------+---------+--------------+-------------
 id     | integer               |           | plain   |              | 
 name   | character varying(25) |           | plain   |              | 
Has OIDs: no
Options: orientation=row, storage_type=USTORE

-- Drop.
openGauss=# DROP TABLE test_alt2;

Examples of Changing a Constraint

Add a NOT NULL constraint to a column.

-- Create a table.
openGauss=# CREATE TABLE test_alt3(pid INT, areaid CHAR(5), name VARCHAR(20));
-- Add a NOT NULL constraint to pid.
openGauss=# ALTER TABLE test_alt3 MODIFY pid NOT NULL;
-- Query.
openGauss=# \d test_alt3
          Table "public.test_alt3"
 Column |         Type          | Modifiers 
--------+-----------------------+-----------
 pid    | integer               | not null
 areaid | character(5)          | 
 name   | character varying(20) |

Cancel the NOT NULL constraint on a column.

openGauss=# ALTER TABLE test_alt3 MODIFY pid NULL;
-- Query.
openGauss=# \d test_alt3
          Table "public.test_alt3"
 Column |         Type          | Modifiers 
--------+-----------------------+-----------
 pid    | integer               | 
 areaid | character(5)          | 
 name   | character varying(20) |

Alter the default value of a column.

-- Alter the default value of id in the test_alt1 table.
openGauss=# ALTER TABLE test_alt3 ALTER COLUMN areaid SET DEFAULT '00000';
-- Query.
openGauss=# \d test_alt3
                 Table "public.test_alt3"
 Column |         Type          |        Modifiers        
--------+-----------------------+-------------------------
 pid    | integer               | 
 areaid | character(5)          | default '00000'::bpchar
 name   | character varying(20) |

-- Drop the default value of id.
openGauss=# ALTER TABLE test_alt3 ALTER COLUMN areaid DROP DEFAULT;
-- Query.
openGauss=# \d test_alt3
          Table "public.test_alt3"
 Column |         Type          | Modifiers 
--------+-----------------------+-----------
 pid    | integer               | 
 areaid | character(5)          | 
 name   | character varying(20) |

Add a table-level constraint.

Directly add a constraint.

-- Add a primary key constraint to the table.
openGauss=# ALTER TABLE test_alt3 ADD CONSTRAINT pk_test3_pid PRIMARY KEY (pid);
-- Query.
openGauss=# \d test_alt3
          Table "public.test_alt3"
 Column |         Type          | Modifiers 
--------+-----------------------+-----------
 pid    | integer               | not null
 areaid | integer               | 
 name   | character varying(20) | 
Indexes:
    "pk_test3_pid" PRIMARY KEY, btree (pid) TABLESPACE pg_default

Create an index and then add a constraint.

-- Create a table.
openGauss=# CREATE TABLE test_alt4(c1 INT, c2 INT);
-- Create an index.
openGauss=# CREATE UNIQUE INDEX pk_test4_c1 ON test_alt4(c1);
-- Associate the created index when adding a constraint.
openGauss=# ALTER TABLE test_alt4 ADD CONSTRAINT pk_test4_c1 PRIMARY KEY USING INDEX pk_test4_c1;
-- Query.
openGauss=# \d test_alt4
   Table "public.test_alt4"
 Column |  Type   | Modifiers 
--------+---------+-----------
 c1     | integer | not null
 c2     | integer | 
Indexes:
    "pk_test4_c1" PRIMARY KEY, btree (c1) TABLESPACE pg_default
-- Drop.
openGauss=# DROP TABLE test_alt4;

Drop a table-level constraint.

-- Drop a constraint.
openGauss=# ALTER TABLE test_alt3 DROP CONSTRAINT IF EXISTS pk_test3_pid;
-- Query.
openGauss=# \d test_alt3
          Table "public.test_alt3"
 Column |         Type          | Modifiers 
--------+-----------------------+-----------
 pid    | integer               | not null
 areaid | integer               | 
 name   | character varying(20) | 

-- Drop.
openGauss=# DROP TABLE test_alt3;