Parameters

Snowflake provides parameters that let you control the behavior of your account, individual user sessions, and objects. All the parameters have default values, which can be set and then overridden at different levels depending on the parameter type (Account, Session, or Object).

In this Topic:

Parameter Hierarchy and Types

This section describes the different types of parameters (Account, Session, and Object) and the levels at which each type can be set.

The following diagram illustrates the hierarchical relationship between the different parameter types and how individual parameters can be overridden at each level:

Hierarchy of account, session, and object parameters

Account Parameters

Account parameters can be set only at the account level by users with the appropriate administrator role. Account parameters are set using the ALTER ACCOUNT command.

Snowflake provides the following account parameters:

Parameter

Notes

ALLOW_ID_TOKEN

Used to enable connection caching in browser-based single sign-on (SSO) for Snowflake-provided clients

CLIENT_ENCRYPTION_KEY_SIZE

Used for encryption of files staged for data loading or unloading; might require additional installation and configuration (see description for details).

INITIAL_REPLICATION_SIZE_LIMIT_IN_TB

NETWORK_POLICY

This is the only account parameter that can be set by either account administrators (ACCOUNTADMIN) or security administrators (SECURITYADMIN).

PERIODIC_DATA_REKEYING

PREVENT_UNLOAD_TO_INLINE_URL

REQUIRE_STORAGE_INTEGRATION_FOR_STAGE_CREATION

REQUIRE_STORAGE_INTEGRATION_FOR_STAGE_OPERATION

SSO_LOGIN_PAGE

Note

By default, account parameters are not displayed in the output of SHOW PARAMETERS. For more information about viewing account parameters, see Viewing the Parameters and Their Values (in this topic).

Session Parameters

Most parameters are session parameters, which can be set at the following levels:

Account

Account administrators can use the ALTER ACCOUNT command to set session parameters for the account. The values set for the account default to individual users and their sessions.

User

Administrators with the appropriate privileges (typically SECURITYADMIN role) can use the ALTER USER command to override session parameters for individual users. The values set for a user default to any sessions started by the user. In addition, users can override default sessions parameters for themselves using ALTER USER.

Session

Users can use the ALTER SESSION to explicitly set session parameters within their sessions.

Note

By default, only session parameters are displayed in the output of SHOW PARAMETERS. For more information about viewing account and object parameters, see Viewing the Parameters and Their Values (in this topic).

Object Parameters

Object parameters can be set at the following levels:

Account

Account administrators can use the ALTER ACCOUNT command to set object parameters for the account. The values set for the account default to the objects created in the account.

Object

Users with the appropriate privileges can use the corresponding CREATE <object> or ALTER <object> commands to override object parameters for an individual object.

Snowflake provides the following object parameters:

Parameter

Object Type

Notes

DATA_RETENTION_TIME_IN_DAYS

Database, Schema, Table

DEFAULT_DDL_COLLATION

Database, Schema, Table

MAX_CONCURRENCY_LEVEL

Warehouse

PIPE_EXECUTION_PAUSED

Schema, Pipe

STATEMENT_QUEUED_TIMEOUT_IN_SECONDS

Warehouse

Also a session parameter (i.e. can be set at both the object and session levels). For inheritance and override details, see the parameter description.

STATEMENT_TIMEOUT_IN_SECONDS

Warehouse

Also a session parameter (i.e. can be set at both the object and session levels). For inheritance and override details, see the parameter description.

NETWORK_POLICY

User

This is the only account parameter that can be set by either account administrators (ACCOUNTADMIN) or security administrators (SECURITYADMIN).

Note

By default, object parameters are not displayed in the output of SHOW PARAMETERS. For more information about viewing object parameters, see Viewing the Parameters and Their Values (in this topic).

Viewing the Parameters and Their Values

Snowflake provides the SHOW PARAMETERS command, which displays a list of the parameters, along with the current and default values for each parameter. The command can be called with different options to determine the type of parameter displayed.

Viewing Session and Object Parameters

By default, the command displays only session parameters:

SHOW PARAMETERS;

To display the object parameters for a specific object, include an IN clause with the object type and name. For example:

SHOW PARAMETERS IN DATABASE mydb;

SHOW PARAMETERS IN WAREHOUSE mywh;

Viewing All Parameters

To display all parameters, including account and object parameters, include an IN ACCOUNT clause:

SHOW PARAMETERS IN ACCOUNT;

Limiting the List of Parameters by Name

The command also supports using a LIKE clause to limit the list of parameters by name. For example:

  • To display the session parameters whose names contain “time”:

    SHOW PARAMETERS LIKE '%time%';
    
  • To display all the parameters whose names start with “time”:

    SHOW PARAMETERS LIKE 'time%' IN ACCOUNT;
    

Note

The LIKE clause must come before the IN clause.

ABORT_DETACHED_QUERY

Type

Session — Can be set for Account » User » Session

Data Type

Boolean

Description

Specifies the action that Snowflake performs for in-progress queries if connectivity is lost due to abrupt termination of a session (e.g. network outage, browser termination, service interruption).

Values

TRUE: In-progress queries are aborted 5 minutes after connectivity is lost.

FALSE: In-progress queries are completed.

Default

FALSE

Note

  • If the user explicitly closes the connection, then all in-progress queries are aborted immediately regardless of the parameter value.

  • Most queries require compute resources to execute. These resources are provided by virtual warehouses, which consume credits while running. With a value of FALSE, if the session terminates, warehouses might continue running and consuming credits to complete any queries that were in progress at the time the session terminated.

ALLOW_ID_TOKEN

Type

Account — Can be set only for Account

Data Type

Boolean

Description

Specifies whether a connection token can be saved in the client-side operating system keystore to promote continuous, secure connectivity without users needing to enter login credentials at the start of each connection attempt to Snowflake. For details and the list of supported Snowflake-provided clients, see (Optional) Using Connection Caching to Minimize the Number of Prompts for Authentication.

Values

TRUE: Stores a connection token in the client-side operating system keystore to enable the client application to perform browser-based SSO without prompting users to authenticate whenever a new connection is established.

FALSE: Does not store a connection token. Users are prompted to authenticate whenever the client application establishes a new connection with Snowflake. SSO to Snowflake is still possible if this parameter is set to false.

Default

FALSE

AUTOCOMMIT

Type

Session — Can be set for Account » User » Session

Data Type

Boolean

Description

Specifies whether autocommit is enabled for the session. Autocommit determines whether a DML statement, when executed without an active transaction, is automatically committed after the statement successfully completes. For more information, see Transactions.

Values

TRUE: Autocommit is enabled.

FALSE: Autocommit is disabled, meaning DML statements must be explicitly committed or rolled back.

Default

TRUE

AUTOCOMMIT_API_SUPPORTED (View-only)

Type

N/A

Data Type

Boolean

Description

For Snowflake internal use only. View-only parameter that indicates whether API support for autocommit is enabled for your account. If the value is TRUE, you can enable or disable autocommit through the APIs for the following drivers/connectors:

BINARY_INPUT_FORMAT

Type

Session — Can be set for Account » User » Session

Data Type

String (Constant)

Description

Input format for binary values.

Values

HEX , BASE64 , or UTF8 / UTF-8

Default

HEX

BINARY_OUTPUT_FORMAT

Type

Session — Can be set for Account » User » Session

Data Type

String (Constant)

Description

Display format for binary values.

Values

HEX or BASE64

Default

HEX

CLIENT_ENABLE_LOG_INFO_STATEMENT_PARAMETERS

Type

Session — Can be set only for Session

Data Type

Boolean

Clients

JDBC

Description

Enables users to log the data values bound to PreparedStatements.

To see the values, you must not only set this session-level parameter to TRUE, but also set the connection parameter named TRACING to either INFO or ALL.

  • Set TRACING to ALL to see all debugging information and all binding information.

  • Set TRACING to INFO to see the binding parameter values and less other debug information.

Values

TRUE or FALSE.

Default

FALSE

CLIENT_ENCRYPTION_KEY_SIZE

Type

Account — Can be set only for Account

Data Type

Integer

Clients

Any

Description

Specifies the AES encryption key size, in bits, used by Snowflake to encrypt/decrypt files stored in internal stages (for loading/unloading data).

Values

128 or 256

Default

128

Note

  • This parameter is not used for encrypting/decrypting files stored in external stages (i.e. S3 buckets or Azure containers). Encryption/decryption of these files is accomplished using an external encryption key explicitly specified in the COPY command or in the named external stage referenced in the command.

  • If you are using the JDBC driver and you wish to set this parameter to 256 (for strong encryption), additional JCE policy files must be installed on each client machine from which data is loaded/unloaded. For more information about installing the required files, see Java Requirements for the JDBC Driver.

  • If you are using the Python connector (or SnowSQL) and you wish to set this parameter to 256 (for strong encryption), no additional installation or configuration tasks are required.

CLIENT_MEMORY_LIMIT

Type

Session — Can be set for Account » User » Session

Data Type

Integer

Clients

JDBC

Description

Parameter that specifies the maximum amount of memory the JDBC driver should use for the result set from queries (in MB). To simplify JVM memory management, the parameter sets a global maximum memory usage limit for all queries.

Note that CLIENT_RESULT_CHUNK_SIZE specifies the maximum size of each set (or chunk) of query results to download (in MB). The driver might require additional memory to process a chunk; if so, it will adjust memory usage during runtime to process at least one thread/query. Verify that CLIENT_MEMORY_LIMIT is set significantly higher than CLIENT_RESULT_CHUNK_SIZE to ensure sufficient memory is available.

Note

  • The driver will attempt to honor the parameter value, but will cap usage at 80% of your system memory.

  • The memory usage limit set in this parameter does not apply to any other JDBC driver operations (e.g. connecting to the database, preparing a query, or PUT and GET statements).

Values

Any valid number of megabytes.

Default

1536 (effectively 1.5 GB)

Most users should not need to set this parameter. If this parameter is not set by the user, the driver starts with the default specified above, but also actively manages its memory conservatively to avoid using up all available memory.

CLIENT_METADATA_REQUEST_USE_CONNECTION_CTX

Type

Session — Can be set for User » Session

Data Type

Boolean

Clients

JDBC, ODBC

Description

For specific ODBC functions and JDBC methods, this parameter can change the default search scope from all databases/schemas to the current database/schema. The narrower search typically returns fewer rows and executes more quickly.

For example, the getTables() JDBC method accepts a database name and schema name as arguments, and returns the names of the tables in the database and schema. If the database and schema arguments are null, then by default, the method searches all databases and all schemas in the account. Setting CLIENT_METADATA_REQUEST_USE_CONNECTION_CTX to TRUE narrows the search to the current database and schema specified by the connection context.

In essence, setting this parameter to TRUE creates the following precedence for database and schema:

  1. Values passed as arguments to the functions/methods.

  2. Values specified in the connection context (if any).

  3. Default (all databases and all schemas).

For more details, see below.

This parameter applies to the following:

  • JDBC driver methods (for the DatabaseMetaData class):

    • getColumns

    • getCrossReference

    • getExportedKeys

    • getForeignKeys

    • getFunctions

    • getImportedKeys

    • getPrimaryKeys

    • getSchemas

    • getTables

  • ODBC driver functions:

    • SQLTables

    • SQLColumns

    • SQLPrimaryKeys

    • SQLForeignKeys

    • SQLGetFunctions

Values

TRUE: If the database and schema arguments are null, then the driver retrieves metadata for only the database and schema specified by the connection context.

The interaction is described in more detail in the table below.

FALSE: If the database and schema arguments are null, then the driver retrieves metadata for all databases and schemas in the account.

Default

FALSE

Additional Notes

The connection context refers to the current database and schema for the session, which can be set using any of the following options:

  1. Specify the default namespace for the user who connects to Snowflake (and initiates the session). This can be set for the user through the CREATE USER or ALTER USER command, but must be set before the user connects.

  2. Specify the database and schema when connecting to Snowflake through the driver.

  3. Issue a USE DATABASE or USE SCHEMA command within the session.

If the database or schema was specified by more than one of these, then the most recent one applies.

When CLIENT_METADATA_REQUEST_USE_CONNECTION_CTX is set to TRUE:

database argument

schema argument

Database used

Schema used

Non-null

Non-null

Argument

Argument

Non-null

Null

Argument

All schemas

Null

Non-null

Connection context

Argument

Null

Null

Connection context

Session context

Note

For the JDBC driver, this behavior applies to version 3.6.27 (and higher). For the ODBC driver, this behavior applies to version 2.12.96 (and higher).

If you want to search only the connection context database, but want to search all schemas within that database, see CLIENT_METADATA_USE_SESSION_DATABASE.

CLIENT_METADATA_USE_SESSION_DATABASE

Type

Session — Can be set for Session

Data Type

Boolean

Clients

JDBC

Description

This parameter applies to only the methods affected by CLIENT_METADATA_REQUEST_USE_CONNECTION_CTX.

This parameter applies only when both of the following conditions are met:

For specific ODBC functions and JDBC methods, this parameter can change the default search scope from all databases to the current database. The narrower search typically returns fewer rows and executes more quickly.

For more details, see below.

Values

TRUE:

The driver searches all schemas in the connection context’s database. (For more details about the connection context, see the documentation for CLIENT_METADATA_REQUEST_USE_CONNECTION_CTX.)

FALSE:

The driver searches all schemas in all databases.

Default

FALSE

Additional Notes

When the database is null and the schema is null and CLIENT_METADATA_REQUEST_USE_CONNECTION_CTX is FALSE:

CLIENT_METADATA_USE_SESSION_DATABASE

Behavior

FALSE

All schemas in all databases are searched.

TRUE

All schemas in the current database are searched.

CLIENT_PREFETCH_THREADS

Type

Session — Can be set for Account » User » Session

Data Type

Integer

Clients

JDBC, Python

Description

Parameter that specifies the number of threads used by the client to pre-fetch large result sets. The driver will attempt to honor the parameter value, but defines the minimum and maximum values (depending on your system’s resources) to improve performance.

Values

1 to 10

Default

4

Most users should not need to set this parameter. If this parameter is not set by the user, the driver starts with the default specified above, but also actively manages its thread count conservatively to avoid using up all available memory.

CLIENT_RESULT_CHUNK_SIZE

Type

Session — Can be set for Account » User » Session

Data Type

Integer

Clients

JDBC

Description

Parameter that specifies the maximum size of each set (or chunk) of query results to download (in MB). The JDBC driver downloads query results in chunks.

Also see CLIENT_MEMORY_LIMIT.

Values

48 to 160

Default

160

Most users should not need to set this parameter. If this parameter is not set by the user, the driver starts with the default specified above, but also actively manages its memory conservatively to avoid using up all available memory.

CLIENT_RESULT_COLUMN_CASE_INSENSITIVE

Type

Session — Can be set for Account » User » Session

Data Type

Boolean

Clients

JDBC

Description

Parameter that indicates whether to match column name case-insensitively in ResultSet.get* methods in JDBC.

Values

TRUE: matches column names case-insensitively.

FALSE: matches column names case-sensitively.

Default

FALSE

CLIENT_SESSION_KEEP_ALIVE

Type

Session — Can be set for Account » User » Session

Data Type

Boolean

Clients

JDBC, ODBC, Python, Node.js

Description

Parameter that indicates whether to force a user to log in again after a period of inactivity in the session.

Values

TRUE: Snowflake keeps the session active indefinitely as long as the connection is active, even if there is no activity from the user.

FALSE: The user must log in again after four hours of inactivity.

Default

FALSE

Note

Currently, the parameter cannot be set at the session level by executing the ALTER SESSION command. For information on setting the parameter at the session level, see the client documentation:

CLIENT_SESSION_KEEP_ALIVE_HEARTBEAT_FREQUENCY

Type

Session — Can be set for Account » User » Session

Data Type

Integer

Clients

SnowSQL, JDBC, Python, Node.js

Description

Number of seconds in-between client attempts to update the token for the session.

Values

900 to 3600

Default

3600

CLIENT_TIMESTAMP_TYPE_MAPPING

Type

Session — Can be set for Account » User » Session

Data Type

String (Constant)

Clients

Any

Description

Specifies the TIMESTAMP_* variation to use when binding timestamp variables for JDBC or ODBC applications that use the bind API to load data.

Values

TIMESTAMP_LTZ or TIMESTAMP_NTZ

Default

TIMESTAMP_LTZ

DATA_RETENTION_TIME_IN_DAYS

Type

Object (for databases, schemas, and tables) — Can be set for Account » Database » Schema » Table

Data Type

Number

Description

Number of days for which Snowflake retains historical data for performing Time Travel actions (SELECT, CLONE, UNDROP) on the object. A value of 0 effectively disables Time Travel for the specified database, schema, or table. For more information, see Understanding & Using Time Travel.

Values

0 or 1 (for Standard Edition)

0 to 90 (for Enterprise Edition or higher)

Default

1

DATE_INPUT_FORMAT

Type

Session — Can be set for Account » User » Session

Data Type

String

Description

Specifies the input format for the DATE data type. For more information, see Date and Time Input / Output.

Values

Any valid, supported date format or AUTO

(AUTO specifies that Snowflake attempts to automatically detect the format of dates stored in the system during the session)

Default

AUTO

DATE_OUTPUT_FORMAT

Type

Session — Can be set for Account » User » Session

Data Type

String

Description

Specifies the display format for the DATE data type. For more information, see Date and Time Input / Output.

Values

Any valid, supported date format

Default

YYYY-MM-DD

DEFAULT_DDL_COLLATION

Type

Object (for databases, schemas, and tables) — Can be set for Account » Database » Schema » Table

Data Type

String

Description

Sets the default collation used for the following DDL operations:

Setting this parameter forces all subsequently-created columns in the affected objects (table, schema, database, or account) to have the specified collation as the default, unless the collation for the column is explicitly defined in the DDL.

For example, if DEFAULT_DDL_COLLATION = 'en-ci', then the following two statements are equivalent:

create table test(c1 integer, c2 string, c3 string collate 'en-cs')

create table test(c1 integer, c2 string collate 'en-ci', c3 string collate 'en-cs');
Values

Any valid, supported collation specification.

Default

Empty string

Note

To set the default collation for the account, use the following command:

The default collation for table columns can be set at the table, schema, or database level during creation or any time afterwards:

ENABLE_UNLOAD_PHYSICAL_TYPE_OPTIMIZATION

Type

Session — Can be set for Account » User » Session

Data Type

Boolean

Description

Specifies whether to set the schema for unloaded Parquet files based on the “logical” column data types (i.e. the types in the unload SQL query or source table) or on the unloaded column values (i.e. the smallest data types and precision that support the values in the output columns of the unload SQL statement or source table).

Values
TRUE: The schema of unloaded Parquet data files is determined by the column values in the unload SQL query or source table. Snowflake optimizes table columns by setting the

smallest precision that accepts all of the values. The unloader follows this pattern when writing values to Parquet files. The data type and precision of an output column are set to the smallest data type and precision that support its values in the unload SQL statement or source table. Accept this setting for better performance and smaller data files.

FALSE: The schema is determined by the logical column data types. Set this value for a consistent output file schema.

Default

TRUE

ERROR_ON_NONDETERMINISTIC_MERGE

Type

Session — Can be set for Account » User » Session

Data Type

Boolean

Description

Specifies whether to return an error when the MERGE command is used to update or delete a target row that joins multiple source rows and the system cannot determine the action to perform on the target row.

Values

TRUE: An error is returned that includes values from one of the target rows that caused the error.

FALSE: No error is returned and the merge completes successfully, but the results of the merge are nondeterministic.

Default

TRUE

ERROR_ON_NONDETERMINISTIC_UPDATE

Type

Session — Can be set for Account » User » Session

Data Type

Boolean

Description

Specifies whether to return an error when the UPDATE command is used to update a target row that joins multiple source rows and the system cannot determine the action to perform on the target row.

Values

TRUE: An error is returned that includes values from one of the target rows that caused the error.

FALSE: No error is returned and the update completes, but the results of the update are nondeterministic.

Default

FALSE

GEOGRAPHY_OUTPUT_FORMAT

Type

Session — Can be set for Account » User » Session

Data Type

String (Constant)

Description

Display format for geospatial values.

For EWKT and EWKB, the SRID is always 4326 in the output. See the note on EWKT and EWKB handling.

Values

GeoJSON, WKT, WKB, EWKT, or EWKB

Default

GeoJSON

INITIAL_REPLICATION_SIZE_LIMIT_IN_TB

Type

Account — Can be set only for Account

Data Type

Number.

Description

Sets the maximum estimated size limit applied when performing the initial replication of a primary database to a secondary database (in TB). This size limit helps prevent accounts from accidentally incurring large database replication charges.

To remove the size limit, set the value to 0.0.

Note that there is currently no default size limit applied to subsequent refreshes of a secondary database.

Values

0.0 and above with a scale of at least 1 (e.g. 20.5, 32.25, 33.333, etc.).

Default

10.0

JDBC_TREAT_DECIMAL_AS_INT

Type

Session — Can be set for Account » User » Session

Data Type

Boolean

Description

Specifies how JDBC processes columns that have a scale of zero (0).

Values

TRUE: JDBC processes a column whose scale is zero as BIGINT.

FALSE: JDBC processes a column whose scale is zero as DECIMAL.

Default

TRUE

JDBC_TREAT_TIMESTAMP_NTZ_AS_UTC

Type

Session — Can be set for Account » User » Session

Data Type

Boolean

Description

Specifies how JDBC processes TIMESTAMP_NTZ values.

By default, when the JDBC driver fetches a value of type TIMESTAMP_NTZ from Snowflake, the driver converts the value from UTC to the current timezone, based on the user’s session.

Users who wish to keep TIMESTAMP_NTZ values in UTC can set this parameter to TRUE.

This parameter applies only to the JDBC driver.

Values

TRUE: The driver does not change the TIMESTAMP_NTZ value; the value stays in UTC.

FALSE: The driver converts the TIMESTAMP_NTZ value from UTC to the user’s

current timezone.

Default

FALSE

JSON_INDENT

Type

Session — Can be set for Account » User » Session

Data Type

Integer

Description

Specifies the number of blank spaces to indent each new element in JSON output in the session. Also specifies whether to insert newline characters after each element.

Values

0 to 16

(a value of 0 returns compact output by removing all blank spaces and newline characters from the output)

Default

2

Note

This parameter does not affect JSON unloaded from a table into a file using the COPY INTO <location> command. The command always unloads JSON data in the ndjson format:

  • Each record from the table separated by a newline character.

  • Within each record, compact formatting (i.e. no spaces or newline characters).

JS_TREAT_INTEGER_AS_BIGINT

Type

Session — Can be set for Session

Data Type

Boolean

Description

Specifies how the Snowflake Node.js Driver processes numeric columns that have a scale of zero (0), for example INTEGER or NUMBER(p, 0).

Values

TRUE: JavaScript processes a column whose scale is zero as Bigint.

FALSE: JavaScript processes a column whose scale is zero as Number.

Default

FALSE

Note

By default, Snowflake INTEGER columns (including BIGINT, NUMBER(p, 0), etc.) are converted to JavaScript’s Number data type. However, the largest legal Snowflake integer values are larger than the largest legal JavaScript Number values. To convert Snowflake INTEGER columns to JavaScript Bigint, which can store larger values than JavaScript Number, set the session parameter JS_TREAT_INTEGER_AS_BIGINT.

For examples of how to use this parameter, see Fetching Integer Data Types as Bigint.

LOCK_TIMEOUT

Type

Session — Can be set for Account » User » Session

Data Type

Number

Description

Number of seconds to wait while trying to lock a resource, before timing out and aborting the statement.

Values

0 to any number (i.e. no limit)

(a value of 0 disables lock waiting, i.e. the statement must acquire the lock immediately or abort; if multiple resources need to be locked by the statement, the timeout applies separately to each lock attempt)

Default

43200 (i.e. 12 hours)

MAX_CONCURRENCY_LEVEL

Type

Object (for warehouses) — Can be set for Account » Warehouse

Data Type

Number

Description

Specifies the concurrency level for SQL statements (i.e. queries and DML) executed by a warehouse cluster. When the level is reached, the operation performed depends on whether the warehouse is single-cluster or multi-cluster:

  • Single-cluster or multi-cluster (in Maximized mode): Statements are queued until already-allocated resources are freed or additional resources are provisioned, which can be accomplished by increasing the size of the warehouse.

  • Multi-cluster (in Auto-scale mode): Additional clusters are started.

MAX_CONCURRENCY_LEVEL can be used in conjunction with the STATEMENT_QUEUED_TIMEOUT_IN_SECONDS parameter to ensure a warehouse is never backlogged.

Note that this parameter does not limit the number of statements that can be executed concurrently by a warehouse cluster. Instead, it serves as an upper-boundary to protect against over-allocation of resources. As each statement is submitted to a warehouse, Snowflake allocates resources for executing the statement; if there aren’t enough resources available, the statement is queued or additional clusters are started, depending on the warehouse.

The actual number of statements executed concurrently by a warehouse cluster might be more or less than the specified level:

  • Smaller, more basic statements: More statements might execute concurrently because small statements generally execute on a subset of the servers in a warehouse cluster. This means they only count as a fraction towards the concurrency level.

  • Larger, more complex statements: Fewer statements might execute concurrently.

Default

8

Tip

This value is a default only and can be changed at any time:

  • Lowering the concurrency level for a warehouse increases the compute resource allocation per statement, which potentially results in faster query performance, particularly for large/complex and multi-statement queries.

  • Raising the concurrency level for a warehouse decreases the compute resource allocation per statement; however, it does not necessarily limit the total number of concurrent queries that can be executed by the warehouse, nor does it necessarily improve total warehouse performance, which depends on the nature of the queries being executed.

Note that, as described earlier, this parameter impacts multi-cluster warehouses (in Auto-scale mode) because Snowflake automatically starts a new cluster within the warehouse to avoid queuing. Thus, lowering the concurrency level for a multi-cluster warehouse (in Auto-scale mode) potentially increases the number of active clusters at any time.

Also, remember that Snowflake automatically allocates resources for each statement when it is submitted and the allocated amount is dictated by the individual requirements of the statement. Based on this, and through observations of user query patterns over time, we’ve selected a default that balances performance and resource usage.

As such, before changing the default, we recommend that you test the change by adjusting the parameter in small increments and observing the impact against a representative set of your queries.

NETWORK_POLICY

Type

Account — Can be set only for Account (can be set by account administrators and security administrators)

Type

Object (for users) — Can be set for Account » User

Data Type

String

Description

Specifies the network policy to enforce for your account. Network policies enable restricting access to your account based on users’ IP address. For more details, see Network Policies.

Values

Any existing network policy (created using CREATE NETWORK POLICY)

Default

None

Note

This is the only account parameter that can be set by security administrators (SECURITYADMIN) or higher.

PERIODIC_DATA_REKEYING

Type

Account — Can be set only for Account

Data Type

Boolean

Description

This parameter only applies to Enterprise Edition (or higher). It enables/disables re-encryption of table data with new keys on a yearly basis to provide additional levels of data protection.

You can enable and disable rekeying at any time. Enabling/disabling rekeying does not result in gaps in your encrypted data:

  • If rekeying is enabled for a period of time and then disabled, all data already tagged for rekeying is rekeyed, but no further data is rekeyed until you re-enable it again.

  • If rekeying is re-enabled, Snowflake automatically rekeys all data that has keys which meet the criteria (i.e. key is older than one year).

For more information about rekeying of encrypted data, see Data Encryption.

Values

TRUE: Data is rekeyed after one year has passed since the data was last encrypted. Rekeying occurs in the background so no down-time is experienced and the affected data/table is always available.

FALSE: Data is not rekeyed.

Default

FALSE

Note

There are charges associated with data rekeying because, after data is rekeyed, the old data (with the previous key encryption) is maintained in Fail-safe for the standard time period (7 days). For this reason, periodic rekeying is disabled by default. To enable periodic rekeying, you must explicitly enable it.

Also, Fail-safe charges for rekeying are not listed individually in your monthly statement; they are included in the Fail-safe total for your account each month.

For more information about Fail-safe, see Understanding & Viewing Fail-safe.

PIPE_EXECUTION_PAUSED

Type

Object — Can be set for Account » Schema » Pipe

Data Type

Boolean

Description

Specifies whether to pause a running pipe, primarily in preparation for transferring ownership of the pipe to a different role:

  • An account administrator (user with the ACCOUNTADMIN role) can set this parameter at the account level, effectively pausing or resuming all pipes in the account.

  • A user with the MODIFY privilege on a schema can pause or resume all pipes in the schema.

  • The pipe owner can set this parameter for a pipe.

Note that setting the parameter at the account or schema level only affects pipes for which the parameter has not already been explicitly set at a lower level (e.g. at the pipe level by the pipe owner).

This enables the practical use case in which an account administrator can pause all pipes at the account level, while a pipe owner can still have an individual pipe running.

Values

TRUE: Pauses the pipe. When the parameter is set to this value, the SYSTEM$PIPE_STATUS function shows the executionState as PAUSED. Note that the pipe owner can continue to submit files to a paused pipe; however, the files are not processed until the pipe is resumed.

FALSE: Resumes the pipe, but only if ownership of the pipe has not been transferred while it was paused. When the parameter is set to this value, the SYSTEM$PIPE_STATUS function shows the executionState as RUNNING.

If ownership of the pipe was transferred to another role after the pipe was paused, this parameter cannot be used to resume the pipe. Instead, use the SYSTEM$PIPE_FORCE_RESUME function to explicitly force the pipe to resume.

This enables the new owner to use SYSTEM$PIPE_STATUS to evaluate the pipe status (e.g. determine how many files are waiting to be loaded) before resuming the pipe.

Default

FALSE (pipes are running by default)

Note

In general, pipes do not need to paused, except for transferring ownership.

PREVENT_UNLOAD_TO_INLINE_URL

Type

Account — Can be set only for Account

Data Type

Boolean

Description

Specifies whether to prevent ad hoc data unload operations to external cloud storage locations (i.e. COPY INTO <location> statements that specify the cloud storage URL and access settings directly in the statement). For an example, see Unloading Data from a Table Directly to Files in an External Location.

Values

TRUE: COPY INTO <location> statements must reference either a named internal (Snowflake) or external stage or an internal user or table stage. A named external stage must store the cloud storage URL and access settings in its definition.

FALSE: Ad hoc data unload operations to external cloud storage locations are permitted.

Default

FALSE

QUERY_TAG

Type

Session — Can be set for Account » User » Session

Data Type

String (up to 2000 characters)

Description

Optional string that can be used to tag queries and other SQL statements executed within a session. The tags are displayed in the output of the QUERY_HISTORY , QUERY_HISTORY_BY_* functions.

Default

None

QUOTED_IDENTIFIERS_IGNORE_CASE

Type

Session — Can be set for Account » User » Session

Table — Can be set for Table » Schema » Database

Data Type

Boolean

Description

Specifies whether the case of letters is ignored for all object identifiers in double quotes.

Values

TRUE: Double-quoted identifiers are case-insensitive (i.e. case is ignored and all letters are treated as uppercase).

FALSE: Double-quoted identifiers are case-sensitive; Snowflake resolves and stores the identifiers in the specified case.

For more information, see Identifier Resolution.

Default

FALSE

For example:

Identifier

Param set to FALSE (default)

Param set to TRUE

"columnname"

resolves to:

columnname

COLUMNNAME

"columnName"

resolves to:

columnName

COLUMNNAME

"ColumnName"

resolves to:

ColumnName

COLUMNNAME

"COLUMNNAME"

resolves to:

COLUMNNAME

COLUMNNAME

REQUIRE_STORAGE_INTEGRATION_FOR_STAGE_CREATION

Type

Account — Can be set only for Account

Data Type

Boolean

Description

Specifies whether to require a storage integration object as cloud credentials when creating a named external stage (using CREATE STAGE) to access a private cloud storage location.

Values

TRUE: Creating an external stage to access a private cloud storage location requires referencing a storage integration object as cloud credentials.

FALSE: Creating an external stage does not require referencing a storage integration object. Users can instead reference explicit cloud provider credentials, such as secret keys or access tokens, if they have been configured for the storage location.

Default

FALSE

REQUIRE_STORAGE_INTEGRATION_FOR_STAGE_OPERATION

Type

Account — Can be set only for Account

Data Type

Boolean

Description

Specifies whether to require using a named external stage that references a storage integration object as cloud credentials when loading data from or unloading data to a private cloud storage location.

Values

TRUE: Loading data from or unloading data to a private cloud storage location requires using a named external stage that references a storage integration object; specifying a named external stage that references explicit cloud provider credentials, such as secret keys or access tokens, produces a user error.

FALSE: Users can load data from or unload data to a private cloud storage location using a named external stage that references explicit cloud provider credentials.

If PREVENT_UNLOAD_TO_INLINE_URL is FALSE, then users can specify the explicit cloud provider credentials directly in the COPY statement.

Default

FALSE

ROWS_PER_RESULTSET

Type

Session — Can be set for Account » User » Session

Data Type

Number

Description

Specifies the maximum number of rows returned in a result set.

Values

0 to any number (i.e. no limit) — a value of 0 specifies no maximum.

Default

0

SAML_IDENTITY_PROVIDER

Type

Account — Can be set only for Account

Data Type

JSON

Description

Enables federated authentication. The parameter parameter accepts a JSON object, enclosed in single quotes, with the following fields:

{
  "certificate": "",
  "ssoUrl": "",
  "type"  : "",
  "label" : ""
}

Where:

certificate

Specifies the certificate (generated by the IdP) that verifies communication between the IdP and Snowflake.

ssoUrl

Specifies the URL endpoint (provided by the IdP) where Snowflake sends the SAML requests.

type

Specifies the type of IdP used for federated authentication ("OKTA" , "ADFS" , "Custom").

label

Specifies the button text for the IdP in the Snowflake login page. The default label is Single Sign On. If you change the default label, the label you specify can only contain alphanumeric characters (i.e. special characters and blank spaces are not currently supported).

Note that, if the "type" field is "Okta", a value for the label field does not need to be specified because Snowflake displays the Okta logo in the button.

For more information, including examples of setting the parameter, see Configuring Snowflake to Use Federated Authentication.

Default

None

SIMULATED_DATA_SHARING_CONSUMER

Type

Session — Can be set for Account » User » Session

Data Type

String

Description

Specifies the name of a consumer account to simulate for testing/validating shared data, particularly shared secure views. When this parameter is set in a session, shared views will return rows as if executed in the specified consumer account rather than the provider account.

Note

At this time, the SIMULATED_DATA_SHARING_CONSUMER session parameter only supports secure views and secure materialized views, but does not support secure UDFs.

For more information, see Introduction to Secure Data Sharing and Working with Shares.

Default

None

Important

This is a session parameter, which means it can be set at the account level; however, it only applies to testing queries on shared views. Because the parameter affects all queries in a session, it should never be set at the account level.

SSO_LOGIN_PAGE

Type

Account — Can be set only for Account

Data Type

Boolean

Description

Disables preview mode for testing SSO (after enabling federated authentication) before rolling it out to users:

Values

TRUE: Preview mode is disabled and users will see the button for Snowflake-initiated SSO for your identity provider (as specified in SAML_IDENTITY_PROVIDER) in the Snowflake main login page.

FALSE: Preview mode is enabled and SSO can be tested using the following URL:

  • If your account is in US West: https://<account_name>.snowflakecomputing.com/console/login?fedpreview=true

  • If your account is in any other region: https://<account_name>.<region_id>.snowflakecomputing.com/console/login?fedpreview=true

For more information, see Configuring Snowflake to Use Federated Authentication.

Default

FALSE

STATEMENT_QUEUED_TIMEOUT_IN_SECONDS

Type

Session and Object (for warehouses)

Can be set for Account » User » Session; can also be set for individual warehouses

Data Type

Number

Description

Amount of time, in seconds, a SQL statement (query, DDL, DML, etc.) remains queued for a warehouse before it is canceled by the system. This parameter can be used in conjunction with the MAX_CONCURRENCY_LEVEL parameter to ensure a warehouse is never backlogged.

The parameter can be set within the session hierarchy. It can also be set for a warehouse to control the queue timeout for all SQL statements processed by the warehouse. When the parameter is set for both a warehouse and a session, the lowest non-zero value is enforced. For example:

  • A warehouse has a queued timeout of 120 seconds.

  • The queued timeout for the session is set to 60 seconds.

The session timeout takes precedence (i.e. any statement submitted in the session is canceled after being queued for longer than 60 seconds).

Values

0 to any number (i.e. no limit) — a value of 0 specifies that no timeout is enforced. A statement will remained queued as long as the queue persists.

Default

0 (i.e. no timeout)

STATEMENT_TIMEOUT_IN_SECONDS

Type

Session and Object (for warehouses)

Can be set for Account » User » Session; can also be set for individual warehouses

Data Type

Number

Description

Amount of time, in seconds, after which a running SQL statement (query, DDL, DML, etc.) is canceled by the system.

The parameter can be set within the session hierarchy. It can also be set for an individual warehouse to control the runtime for all SQL statements processed by the warehouse. When the parameter is set for both a warehouse and a session, the lowest non-zero value is enforced. For example:

  • A warehouse has a timeout of 1000 seconds.

  • The timeout for the session is set to 500 seconds.

The session timeout takes precedence (i.e. any statement submitted in the session is canceled after running for longer than 500 seconds).

Values

0 to 604800 (i.e. 7 days) — a value of 0 specifies that the maximum timeout value is enforced.

Default

172800 (i.e. 2 days)

STRICT_JSON_OUTPUT

Type

Session — Can be set for Account » User » Session

Data Type

Boolean

Description

This parameter specifies whether JSON output in a session is compatible with the general standard (as described by http://json.org).

By design, Snowflake allows JSON input that contains non-standard values; however, these non-standard values might result in Snowflake outputting JSON that is incompatible with other platforms and languages. This parameter, when enabled, ensures that Snowflake outputs valid/compatible JSON.

Values

TRUE: Strict JSON output is enabled, enforcing the following behavior:

  • Missing and undefined values in input mapped to JSON NULL.

  • Non-finite numeric values in input (Infinity, -Infinity, NaN, etc.) mapped to strings with valid JavaScript representations. This enables compatibility with JavaScript and also allows conversion of these values back to numeric values.

FALSE: Strict JSON output is not enabled.

Default

FALSE

For example:

Non-standard JSON Input

Param set to FALSE (default)

Param set to TRUE

[289, 2188,]

outputs:

[ 289, 2188, undefined ]

[ 289, 2188, null ]

[undefined, undefined]

outputs:

[ undefined, undefined ]

[ null, null ]

[Infinity,inf,-Infinity,-inf]

outputs:

[ Infinity, Infinity, -Infinity, -Infinity ]

[ "Infinity", "Infinity", "-Infinity", "-Infinity" ]

[NaN,nan]

outputs:

[ NaN, NaN ]

[ "NaN", "NaN" ]

TIMESTAMP_DAY_IS_ALWAYS_24H

Type

Session — Can be set for Account » User » Session

Data Type

Boolean

Description

Specifies whether the DATEADD function (and its aliases) always consider a day to be exactly 24 hours for expressions that span multiple days.

Values

TRUE: A day is always exactly 24 hours.

FALSE: A day is not always 24 hours.

Default

FALSE

Important

If set to TRUE, the actual time of day might not be preserved when daylight saving time (DST) is in effect. For example:

alter session set TIMESTAMP_DAY_IS_ALWAYS_24H = true;

-- With DST beginning on 2018-03-11 at 2 AM, America/Los_Angeles time zone
select dateadd(day, 1, '2018-03-10 09:00:00'::TIMESTAMP_LTZ), dateadd(day, 1, '2018-11-03 09:00:00'::TIMESTAMP_LTZ);

+-------------------------------------------------------+-------------------------------------------------------+
| DATEADD(DAY, 1, '2018-03-10 09:00:00'::TIMESTAMP_LTZ) | DATEADD(DAY, 1, '2018-11-03 09:00:00'::TIMESTAMP_LTZ) |
|-------------------------------------------------------+-------------------------------------------------------|
| 2018-03-11 10:00:00.000 -0700                         | 2018-11-04 08:00:00.000 -0800                         |
+-------------------------------------------------------+-------------------------------------------------------+

alter session set TIMESTAMP_DAY_IS_ALWAYS_24H = false;

select dateadd(day, 1, '2018-03-10 09:00:00'::TIMESTAMP_LTZ), dateadd(day, 1, '2018-11-03 09:00:00'::TIMESTAMP_LTZ);

+-------------------------------------------------------+-------------------------------------------------------+
| DATEADD(DAY, 1, '2018-03-10 09:00:00'::TIMESTAMP_LTZ) | DATEADD(DAY, 1, '2018-11-03 09:00:00'::TIMESTAMP_LTZ) |
|-------------------------------------------------------+-------------------------------------------------------|
| 2018-03-11 09:00:00.000 -0700                         | 2018-11-04 09:00:00.000 -0800                         |
+-------------------------------------------------------+-------------------------------------------------------+

TIMESTAMP_INPUT_FORMAT

Type

Session — Can be set for Account » User » Session

Data Type

String

Description

Specifies the input format for the TIMESTAMP data type alias. For more information, see Date and Time Input / Output.

Values

Any valid, supported timestamp format or AUTO

(AUTO specifies that Snowflake attempts to automatically detect the format of timestamps stored in the system during the session)

Default

AUTO

TIMESTAMP_LTZ_OUTPUT_FORMAT

Type

Session — Can be set for Account » User » Session

Data Type

String

Description

Specifies the display format for the TIMESTAMP_LTZ data type. If no format is specified, defaults to TIMESTAMP_OUTPUT_FORMAT. For more information, see Date and Time Input / Output.

Values

Any valid, supported timestamp format

Default

None

TIMESTAMP_NTZ_OUTPUT_FORMAT

Type

Session — Can be set for Account » User » Session

Data Type

String

Description

Specifies the display format for the TIMESTAMP_NTZ data type.

Values

Any valid, supported timestamp format

Default

YYYY-MM-DD HH24:MI:SS.FF3

TIMESTAMP_OUTPUT_FORMAT

Type

Session — Can be set for Account » User » Session

Data Type

String

Description

Specifies the display format for the TIMESTAMP data type alias. For more information, see Date and Time Input / Output.

Values

Any valid, supported timestamp format

Default

YYYY-MM-DD HH24:MI:SS.FF3 TZHTZM

TIMESTAMP_TYPE_MAPPING

Type

Session — Can be set for Account » User » Session

Data Type

String

Description

Specifies the TIMESTAMP_* variation that the TIMESTAMP data type alias maps to.

Values

TIMESTAMP_LTZ , TIMESTAMP_NTZ , or TIMESTAMP_TZ

Default

TIMESTAMP_NTZ

TIMESTAMP_TZ_OUTPUT_FORMAT

Type

Session — Can be set for Account » User » Session

Data Type

String

Description

Specifies the display format for the TIMESTAMP_TZ data type. If no format is specified, defaults to TIMESTAMP_OUTPUT_FORMAT. For more information, see Date and Time Input / Output.

Values

Any valid, supported timestamp format

Default

None

TIMEZONE

Type

Session — Can be set for Account » User » Session

Data Type

String (Constant)

Description

Specifies the time zone for the session. For a list of time zones, see the list of tz database time zones (in Wikipedia).

Values

Standard iana.org time zones:

America/Los_Angeles , Europe/London , UTC , Etc/GMT , etc.

Default

America/Los_Angeles

Note

Time zone names are case-sensitive and must be enclosed in single quotes (e.g. 'UTC').

Also, certain simple time zones, such as PDT, are not currently supported.

TIME_INPUT_FORMAT

Type

Session — Can be set for Account » User » Session

Data Type

String

Description

Specifies the input format for the TIME data type. For more information, see Date and Time Input / Output.

Values

Any valid, supported time format or AUTO

(AUTO specifies that Snowflake attempts to automatically detect the format of times stored in the system during the session)

Default

AUTO

TIME_OUTPUT_FORMAT

Type

Session — Can be set for Account » User » Session

Data Type

String

Description

Specifies the display format for the TIME data type. For more information, see Date and Time Input / Output.

Values

Any valid, supported time format

Default

HH24:MI:SS

TRANSACTION_ABORT_ON_ERROR

Type

Session — Can be set for Account » User » Session

Data Type

BOOLEAN

Description

Specifies the action to perform when a statement issued within a non-autocommit transaction returns with an error.

Values

TRUE: The non-autocommit transaction is aborted. All statements issued inside that transaction will fail until a commit or rollback statement is executed to close that transaction.

FALSE: The non-autocommit transaction is not aborted.

Default

FALSE

TRANSACTION_DEFAULT_ISOLATION_LEVEL

Type

Session — Can be set for Account » User » Session

Data Type

String

Description

Specifies the isolation level for transactions in the user session.

Values

READ COMMITTED (only currently-supported value)

Default

READ COMMITTED

TWO_DIGIT_CENTURY_START

Type

Session — Can be set for Account » User » Session

Data Type

Number

Description

Specifies the “century start” year for 2-digit years (i.e. the earliest year such dates can represent). This parameter prevents ambiguous dates when importing or converting data with the YY date format component (i.e. years represented as 2 digits).

Values

1900 to 2100 (any value outside of this range returns an error)

Default

1970

For example:

Year

Param set to 1900

Param set to 1970 (default)

Param set to 1980

Param set to 1990

Param set to 2000

00

becomes:

1900

2000

2000

2000

2000

79

becomes:

1979

1979

2079

2079

2079

89

becomes:

1989

1989

1989

2089

2089

99

becomes:

1999

1999

1999

1999

2099

UNSUPPORTED_DDL_ACTION

Type

Session — Can be set for Account » User » Session

Data Type

String (Constant)

Description

Specifies whether an unsupported (i.e. non-default) value specified for a constraint property returns an error.

Values

IGNORE: Snowflake does not return an error for unsupported values.

FAIL: Snowflake returns an error for unsupported values.

Default

IGNORE

Important

This parameter does not determine whether the constraint is created. Snowflake does not create constraints using unsupported values, regardless of how this parameter is set.

For more information, see Constraint Properties.

USE_CACHED_RESULT

Type

Session — Can be set for Account » User » Session

Data Type

Boolean

Description

Specifies whether to reuse cached query results, if available, when a matching query is submitted.

Values

TRUE: When a query is submitted, Snowflake checks for matching query results for previously-executed queries and, if a matching result exists, uses the result instead of executing the query. This can help reduce query time because Snowflake retrieves the result directly from the cache.

FALSE: Snowflake executes each query when submitted, regardless of whether a matching query result exists.

Default

TRUE

WEEK_OF_YEAR_POLICY

Type

Session — Can be set for Account » User » Session

Data Type

Number

Description

Specifies how the weeks in a given year are computed.

Values

0: The semantics used are equivalent to the ISO semantics, in which a week belongs to a given year if at least 4 days of that week are in that year.

1: January 1 is included in the first week of the year and December 31 is included in the last week of the year.

Default

0 (i.e. ISO-like behavior)

Tip

1 is the most common value, based on feedback we’ve received. For more information, including examples, see Calendar Weeks and Weekdays.

WEEK_START

Type

Session — Can be set for Account » User » Session

Data Type

Number

Description

Specifies the first day of the week (used by week-related date functions).

Values

0: Legacy Snowflake behavior is used (i.e. ISO-like semantics).

1 (Monday) to 7 (Sunday): All the week-related functions use weeks that start on the specified day of the week.

Default

0 (i.e. legacy Snowflake behavior)

Tip

1 is the most common value, based on feedback we’ve received. For more information, including examples, see Calendar Weeks and Weekdays.