CREATE WAREHOUSE¶

Syntax¶

CREATE [ OR REPLACE ] WAREHOUSE [ IF NOT EXISTS ] <name>
       [ [ WITH ] objectProperties ]
       [ [ WITH ] TAG ( <tag_name> = '<tag_value>' [ , <tag_name> = '<tag_value>' , ... ] ) ]
       [ objectParams ]

Where:

objectProperties ::=
  WAREHOUSE_TYPE = { STANDARD | 'SNOWPARK-OPTIMIZED' }
  WAREHOUSE_SIZE = { XSMALL | SMALL | MEDIUM | LARGE | XLARGE | XXLARGE | XXXLARGE | X4LARGE | X5LARGE | X6LARGE }
  RESOURCE_CONSTRAINT = { STANDARD_GEN_1 | STANDARD_GEN_2 | MEMORY_1X | MEMORY_1X_x86 | MEMORY_16X | MEMORY_16X_x86 | MEMORY_64X | MEMORY_64X_x86 }
  MAX_CLUSTER_COUNT = <num>
  MIN_CLUSTER_COUNT = <num>
  SCALING_POLICY = { STANDARD | ECONOMY }
  AUTO_SUSPEND = { <num> | NULL }
  AUTO_RESUME = { TRUE | FALSE }
  INITIALLY_SUSPENDED = { TRUE | FALSE }
  RESOURCE_MONITOR = <monitor_name>
  COMMENT = '<string_literal>'
  ENABLE_QUERY_ACCELERATION = { TRUE | FALSE }
  QUERY_ACCELERATION_MAX_SCALE_FACTOR = <num>

Copy

objectParams ::=
  MAX_CONCURRENCY_LEVEL = <num>
  STATEMENT_QUEUED_TIMEOUT_IN_SECONDS = <num>
  STATEMENT_TIMEOUT_IN_SECONDS = <num>

Copy

Variant syntax¶

CREATE OR ALTER WAREHOUSE <name>
     [ [ WITH ] objectProperties ]
     [ objectParams ]

objectProperties ::=
  WAREHOUSE_TYPE = { STANDARD | 'SNOWPARK-OPTIMIZED' }
  WAREHOUSE_SIZE = { XSMALL | SMALL | MEDIUM | LARGE | XLARGE | XXLARGE | XXXLARGE | X4LARGE | X5LARGE | X6LARGE }
  RESOURCE_CONSTRAINT = { STANDARD_GEN_1 | STANDARD_GEN_2 | MEMORY_1X | MEMORY_1X_x86 | MEMORY_16X | MEMORY_16X_x86 | MEMORY_64X | MEMORY_64X_x86 }
  MAX_CLUSTER_COUNT = <num>
  MIN_CLUSTER_COUNT = <num>
  SCALING_POLICY = { STANDARD | ECONOMY }
  AUTO_SUSPEND = { <num> | NULL }
  AUTO_RESUME = { TRUE | FALSE }
  INITIALLY_SUSPENDED = { TRUE | FALSE }
  RESOURCE_MONITOR = <monitor_name>
  COMMENT = '<string_literal>'
  ENABLE_QUERY_ACCELERATION = { TRUE | FALSE }
  QUERY_ACCELERATION_MAX_SCALE_FACTOR = <num>

objectParams ::=
  MAX_CONCURRENCY_LEVEL = <num>
  STATEMENT_QUEUED_TIMEOUT_IN_SECONDS = <num>
  STATEMENT_TIMEOUT_IN_SECONDS = <num>

Required parameters¶

name

Identifier for the virtual warehouse; must be unique for your account.

In addition, the identifier must start with an alphabetic character and cannot contain spaces or special characters unless the entire identifier string is enclosed in double quotes (for example, "My object"). Identifiers enclosed in double quotes are also case-sensitive.

For more details, see Identifier requirements.

Optional properties (objectProperties)¶

WAREHOUSE_TYPE = { STANDARD | 'SNOWPARK-OPTIMIZED' }

Specifies the warehouse type.

Valid values:

• STANDARD, 'STANDARD'

• 'SNOWPARK-OPTIMIZED'

Default:

STANDARD

Note

To use a value that contains a hyphen ('SNOWPARK-OPTIMIZED'), you must enclose the value in single quotes, as shown.

WAREHOUSE_SIZE = { XSMALL | SMALL | MEDIUM | LARGE | XLARGE | XXLARGE | XXXLARGE | X4LARGE | X5LARGE | X6LARGE }

Specifies the size of the virtual warehouse. The size determines the amount of compute resources in each cluster in the warehouse and, therefore, the number of credits consumed while the warehouse is running.

Valid values:

Supported Values	Synonyms
XSMALL	'X-SMALL'
SMALL
MEDIUM
LARGE
XLARGE	'X-LARGE'
XXLARGE	X2LARGE , '2X-LARGE'
XXXLARGE	X3LARGE , '3X-LARGE'
X4LARGE	'4X-LARGE'
X5LARGE	'5X-LARGE'
X6LARGE	'6X-LARGE'

Default:

XSMALL

Note

• X5LARGE and X6LARGE sizes for Snowpark-optimized warehouses are only supported with the MEMORY_16X resource constraint.

• X5LARGE and X6LARGE sizes aren’t supported for standard warehouses that use the STANDARD_GEN_2 resource constraint.

• The default size for Snowpark-optimized warehouses is MEDIUM.

• To use a value that contains a hyphen (for example, '2X-LARGE'), you must enclose the value in single quotes, as shown.

• Larger warehouse sizes 5X-Large and 6X-Large are generally available in all Amazon Web Services (AWS) and Microsoft Azure regions.

  Larger warehouse sizes are in preview in US Government regions (requires FIPS support on ARM).

RESOURCE_CONSTRAINT = { STANDARD_GEN_1 | STANDARD_GEN_2 | MEMORY_1X| MEMORY_1X_x86 | MEMORY_16X | MEMORY_16X_x86 | MEMORY_64X | MEMORY_64X_x86 }

Preview Feature — Open

The 1 TB resource constraints (MEMORY_64X and MEMORY_64X_x86) are available as a preview feature. The 1 TB constraints are available only on the Amazon Web Services (AWS) cloud platform.

All other MEMORY_* resource constraint sizes are generally available and are available for all cloud platforms.

Specifies the memory and CPU architecture for Snowpark-optimized warehouses, or generation 1 or generation 2 capabilities for standard warehouses.

The following table includes the valid values for the property, available memory, CPU architecture, and the minimum warehouse size required for the resource_constraint setting. For more information about regions and cloud service providers where generation 2 standard warehouses are available, see Snowflake generation 2 standard warehouses.

Valid values:

Value	Memory (up to)	CPU architecture	Min warehouse size required	Max warehouse size
STANDARD_GEN_1	16 GB	Standard	XSMALL	X6LARGE
STANDARD_GEN_2	16 GB	Standard (generation 2)	XSMALL	X4LARGE
MEMORY_1X	16 GB	Standard	XSMALL	X4LARGE
MEMORY_1X_x86	16 GB	x86	XSMALL	X4LARGE
MEMORY_16X	256 GB	Standard	MEDIUM	X6LARGE
MEMORY_16X_x86	256 GB	x86	MEDIUM	X4LARGE
MEMORY_64X	1 TB	Standard	LARGE	X4LARGE
MEMORY_64X_x86	1 TB	x86	LARGE	X4LARGE

Default value:

MEMORY_16X for Snowpark-optimized warehouses. STANDARD_GEN_1 for standard warehouses.

MAX_CLUSTER_COUNT = num

Specifies the maximum number of clusters for a multi-cluster warehouse. For a single-cluster warehouse, this value is always 1.

Valid values:

1 to an upper limit that varies depending on warehouse size.

Note that specifying a value greater than 1 indicates the warehouse is a multi-cluster warehouse; however, the value can only be set to a higher value in Snowflake Enterprise Edition (or higher).

For more information, including the upper limit for each warehouse size, see Multi-cluster warehouses.

Default:

1 (single-cluster warehouse)

Tip

For Snowflake Enterprise Edition (or higher), we recommend always setting the value greater than 1 to help maintain high-availability and optimal performance of a multi-cluster warehouse. This also helps ensure continuity in the unlikely event that a cluster fails.

MIN_CLUSTER_COUNT = num

Specifies the minimum number of clusters for a multi-cluster warehouse (only applies to multi-cluster warehouses).

Valid values:

1 to the value of MAX_CLUSTER_COUNT. The upper limit for MAX_CLUSTER_COUNT varies depending on the warehouse size.

MIN_CLUSTER_COUNT must be equal to or less than MAX_CLUSTER_COUNT:

• If both parameters are equal, the warehouse runs in Maximized mode.

• If MIN_CLUSTER_COUNT is less than MAX_CLUSTER_COUNT, the warehouse runs in Auto-scale mode.

For more information, including the upper limit for each warehouse size, see Multi-cluster warehouses.

Default:

1

SCALING_POLICY = { STANDARD | ECONOMY }

Specifies the policy for automatically starting and shutting down clusters in a multi-cluster warehouse running in Auto-scale mode.

Valid values:

• STANDARD: Minimizes queuing by starting clusters.

• ECONOMY: Conserves credits by favoring keeping running clusters fully-loaded.

For a more detailed description, see Setting the scaling policy for a multi-cluster warehouse.

Default:

STANDARD

AUTO_SUSPEND = { num | NULL }

Specifies the number of seconds of inactivity after which a warehouse is automatically suspended.

Valid values:

Any integer 0 or greater, or NULL:

• The background process that suspends a warehouse runs approximately every 30 seconds and therefore, the setting for this property is not intended for enabling precise control over warehouse suspension.

• Setting a value less than 30, or a value that is not a multiple of 30, is allowed but might not result in the expected behavior due to the 30 second poll interval for warehouse suspension.

• Setting a 0 or NULL value means the warehouse never suspends.

Default:

600 (the warehouse suspends automatically after 10 minutes of inactivity)

Important

Setting AUTO_SUSPEND to 0 or NULL is not recommended, unless your query workloads require a continually running warehouse. Note that this can result in significant consumption of credits (and corresponding charges), particularly for larger warehouses.

AUTO_RESUME = { TRUE | FALSE }

Specifies whether to automatically resume a warehouse when a SQL statement (for example, query) is submitted to it.

Valid values:

• TRUE: The warehouse resumes when a new query is submitted.

• FALSE: The warehouse only resumes when explicitly resumed using ALTER WAREHOUSE or through the Snowflake web interface.

Default:

TRUE (the warehouse resumes automatically when a SQL statement is submitted to it)

INITIALLY_SUSPENDED = { TRUE | FALSE }

Specifies whether the warehouse is created initially in the ‘Suspended’ state.

Valid values:

• TRUE: The warehouse is created, but suspended.

• FALSE: The warehouse starts running after it is created.

Default:

FALSE

RESOURCE_MONITOR = monitor_name

Specifies the name of a resource monitor that is explicitly assigned to the warehouse. When a resource monitor is explicitly assigned to a warehouse, the monitor controls the monthly credits used by the warehouse (and all other warehouses to which the monitor is assigned).

Valid values:

Any existing resource monitor.

For more details, see Working with resource monitors.

Default:

No value (no resource monitor assigned to the warehouse)

Tip

To view all resource monitors and their identifiers, use the SHOW RESOURCE MONITORS command.

COMMENT = 'string_literal'

Specifies a comment for the warehouse.

TAG ( tag_name = 'tag_value' [ , tag_name = 'tag_value' , ... ] )

Specifies the tag name and the tag string value.

The tag value is always a string, and the maximum number of characters for the tag value is 256.

For information about specifying tags in a statement, see Tag quotas for objects and columns.

Optional parameters (objectParams)¶

MAX_CONCURRENCY_LEVEL = num

Object parameter that specifies the concurrency level for SQL statements (i.e. queries and DML) executed by a warehouse cluster.

For a detailed description of this parameter, see MAX_CONCURRENCY_LEVEL.

STATEMENT_QUEUED_TIMEOUT_IN_SECONDS = num

Object parameter that specifies the time, in seconds, a SQL statement (query, DDL, DML, etc.) can be queued on a warehouse before it is canceled by the system.

For a detailed description of this parameter, see STATEMENT_QUEUED_TIMEOUT_IN_SECONDS.

STATEMENT_TIMEOUT_IN_SECONDS = num

Object parameter that specifies the time, in seconds, after which a running SQL statement (query, DDL, DML, etc.) is canceled by the system.

For a detailed description of this parameter, see STATEMENT_TIMEOUT_IN_SECONDS.

Access control requirements¶

A role used to execute this operation must have the following privileges at a minimum:

For instructions on creating a custom role with a specified set of privileges, see Creating custom roles.

For general information about roles and privilege grants for performing SQL actions on securable objects, see Overview of Access Control.

General usage notes¶

• Creating a virtual warehouse automatically sets it as the warehouse in use for the current session (equivalent to using the USE WAREHOUSE command for the warehouse).

  To change the warehouse in use for the current session, execute an explicit USE WAREHOUSE statement after the CREATE WAREHOUSE statement. For example, create warehouse my_wh but continue to use the current warehouse, not my_wh, to execute additional statements:

  SET current_wh_name = (SELECT CURRENT_WAREHOUSE());
  
  CREATE OR REPLACE WAREHOUSE my_wh
    WAREHOUSE_SIZE = 'XSMALL';
  
  USE WAREHOUSE IDENTIFIER($current_wh_name);
  

  Copy

• Regarding metadata:

  Attention

  Customers should ensure that no personal data (other than for a User object), sensitive data, export-controlled data, or other regulated data is entered as metadata when using the Snowflake service. For more information, see Metadata fields in Snowflake.

• Using OR REPLACE is the equivalent of using DROP WAREHOUSE on the existing warehouse and then creating a new warehouse with the same name.

  CREATE OR REPLACE <object> statements are atomic. That is, when an object is replaced, the old object is deleted and the new object is created in a single transaction.

  Any queries running on the dropped warehouse are aborted.

• The OR REPLACE and IF NOT EXISTS clauses are mutually exclusive. They can’t both be used in the same statement.

• Initial creation and resumption of a Snowpark-optimized virtual warehouse may take longer than standard warehouses.

CREATE OR ALTER WAREHOUSE usage notes¶

Limitations

• All limitations of the ALTER WAREHOUSE command apply.

• The INITIALLY_SUSPENDED property cannot be altered (SET or UNSET).

Warehouse parameters and properties

• The absence of a property or parameter that was previously set in the modified warehouse definition results in unsetting it.

• Unsetting an explicit parameter value results in setting it to the default parameter value.

Data governance

• Setting or unsetting a tag or policy on a warehouse using a CREATE OR ALTER WAREHOUSE statement is not supported.

• Existing policies or tags cannot be not altered by a CREATE OR ALTER WAREHOUSE statement and remain unchanged.

Billing and Pricing¶

For information on Snowpark-optimized warehouse credit consumption, see Table 1(a): Snowflake Credit Table for Virtual Warehouse Services in the Snowflake Service Consumption Table.

Examples¶

CREATE WAREHOUSE so_warehouse WITH
  WAREHOUSE_TYPE = 'SNOWPARK-OPTIMIZED'
  WAREHOUSE_SIZE = xlarge
  RESOURCE_CONSTRAINT = 'MEMORY_16X_x86';

CREATE OR ALTER WAREHOUSE so_warehouse
  WAREHOUSE_TYPE = 'SNOWPARK_OPTIMIZED'
  WAREHOUSE_SIZE = 'X-LARGE'
  RESOURCE_CONSTRAINT = 'MEMORY_16X_X86'
  AUTO_RESUME = TRUE
  COMMENT = 'Snowpark warehouse for ingestion';

CREATE OR ALTER WAREHOUSE so_warehouse
  WAREHOUSE_TYPE = 'SNOWPARK-OPTIMIZED'
  WAREHOUSE_SIZE = 'X-LARGE'
  RESOURCE_CONSTRAINT = 'MEMORY_16X_X86'
  AUTO_RESUME = FALSE
  COMMENT = 'Snowpark warehouse for ingestion (disabled for auto-resume)';