Snowflake Data Warehouse

You are looking at an older version of the documentation. The latest version is found here.

Snowflake is a cloud-based data storage and analytics service providing relational database support for both structured data, such as CSV files and tables, and semi-structured data, including JSON, Avro, Parquet, etc. It can be used both as data source and as analytical storage.

Connector Configuration

Snowflake requires the following parameters:

Parameter	Description
`host`	A Snowflake endpoint based on account name
`user-name`	Name of the user's account (provided by Snowflake)
`password`	Password for the specified user
`db`	Default database to use once connected
`warehouse`	Virtual warehouse to use once connected

Example

SQL

CALL SYSADMIN.createConnection(name => 'snowflake', jbossCLITemplateName => 'snowflake', connectionOrResourceAdapterProperties => 'host=<your_account_name.snowflakecomputing.com,db=<database>,user-name=<user_name>,password=<user_password>,warehouse=<warehouse>') ;;

CALL SYSADMIN.createDataSource(name => 'snowflake', translator => 'snowflake', modelProperties => 'importer.defaultSchema=PUBLIC,importer.schemaPattern=PUBLIC,importer.useFullSchemaName=FALSE,importer.tableTypes="TABLE,VIEW"', translatorProperties => supportsNativeQueries=true') ;;

Translator Properties

Amazon AWS S3 Upload

Inserting data into Snowflake using the standard JDBC protocol can be extremely slow, so we recommend configuring and using Amazon AWS S3 (S3LOAD) for productive usage.

To configure S3LOAD, you will need the following translator parameters:

Parameter	Description
`uploadMode=s3load`	Explicitly specifies the S3LOAD mode
`region`	AWS S3 region endpoint
`bucketName`	Optional parameter; bucket name to upload data files to. It is not necessarily an already existing bucket: you can use the `createBucket` parameter to create a new bucket or the `bucketPrefix` parameter to use a temporary bucket (see the next two rows)
`createBucket`	Optional parameter; specifies if the bucket set in the `bucketName` parameter should be created if it does not exist. The default value is `FALSE`
`bucketPrefix`	Optional parameter; prefix of the temporary bucket to upload data files to if `bucketName` is not specified; must comply with Amazon S3 bucket naming rules (note: 36 characters would be added to the bucket prefix when creating a temporary bucket)
`keyId`	AWS S3 key ID
`secretKey`	AWS S3 secret key

Examples

1. Snowflake as Data Source

SQL

CALL SYSADMIN.createConnection(name => 'snowflake', jbossCLITemplateName => 'snowflake', connectionOrResourceAdapterProperties => 'host=<your_account_name.snowflakecomputing.com,db=<database>,user-name=<user_name>,password=<user_password>,warehouse=<warehouse>') ;;

CALL SYSADMIN.createDataSource(name => 'snowflake', translator => 'snowflake', modelProperties => 'importer.defaultSchema=PUBLIC,importer.schemaPattern=PUBLIC,importer.useFullSchemaName=FALSE,importer.tableTypes="TABLE,VIEW"', translatorProperties => 'supportsNativeQueries=true,uploadMode=s3Load,region=<region>,bucketName=<bucket_name>,createBucket=true,keyId=<key_ID>,secretKey="<secret_key>"') ;;

2. Snowflake as Analytical Storage

SQL

CALL SYSADMIN.createConnection(name => 'snowflake', jbossCLITemplateName => 'snowflake', connectionOrResourceAdapterProperties => 'host=<your_account_name.snowflakecomputing.com,db=<database>,user-name=<user_name>,password=<user_password>,warehouse=<warehouse>') ;;

CALL SYSADMIN.createDataSource(name => 'snowflake', translator => 'snowflake', modelProperties => 'importer.defaultSchema=PUBLIC,importer.schemaPattern=PUBLIC,importer.useFullSchemaName=FALSE,importer.tableTypes="TABLE,VIEW"', translatorProperties => 'supportsNativeQueries=true,uploadMode=s3Load,region=<region>,bucketName=<bucket_name>,createBucket=true,keyId=<key_ID>,secretKey="<secret_key>"') ;;

CALL SYSADMIN.setCurrentDWH('dwh', 'PUBLIC') ;;

To read data from an S3 bucket, the security and access management policies on the bucket must allow Snowflake to access the bucket. There must be an IAM user with the required permissions to access the specified S3 bucket. This one-time setup involves establishing access permissions on a bucket and associating the required permissions with an IAM user. Access management instructions can be found in Configuring AWS IAM User Credentials.

Proxy settings for Amazon AWS S3 upload can be configured using System Properties.

Microsoft Azure Blob Storage Upload

As an alternative to Amazon AWS S3, you can configure data upload via Microsoft Azure Blob Storage.

Here are the translator parameters you will need:

Parameter	Description
`uploadMode`	Microsoft Azure Blob Storage upload mode: `blob_azure`, `blob_azure_direct_write`, or `blob_azure_multi_thread`
`storageAccountName`	Microsoft Azure Blob Storage account name
`storageAccountKey`	Microsoft Azure Blob Storage account key
`containerName`	Name of the container defined in the Microsoft Azure Blob Storage account. This property is case-sensitive as the container name is also case-sensitive in Microsoft Azure Blob Storage
`defaultEndpointsProtocol`	Endpoint protocol
`tempFolder`	Path to a temporary files folder
`keepTempFiles`	If set to `TRUE`, temporary files will not be deleted after uploading
`uploadZipped`	If set to `TRUE`, zipped files will be uploaded
`azureKey`	Key for encrypting data while it is stored in the Azure Blob Storage, must be a base64-encoded 128, 192, or 256-bit value. When set to an empty value (i.e. `azureKey=`), the data does not get encrypted. The default behaviour depends on the way the data source connects to the Azure Blob Storage (see below)

To set up secure access to data files in a Microsoft Azure Blob container, you can use one of two options: creating a SAS token or configuring a Snowflake Storage integration.

Connecting via SAS Token

To use the SAS token, configure the sasToken property with the SAS token used to access the Microsoft Azure Blob Storage from Snowflake with the following parameters:

Parameter	Description
`sasToken`	SAS token used to access the Microsoft Azure Blob Storage from Snowflake
`createContainer`	Specifies if the container set in the `containerName` parameter should be created if it does not exist; optional. Default: `FALSE`

Please keep in mind the following:

The containerName parameter specifies a temporary Azure Blob Storage container to be used (created and deleted for each transaction) if this parameter is not set;
By default, when azureKey is not set, the data is encrypted with a key generated by the server.

Example

SQL

CALL SYSADMIN.createConnection(name => 'snowflake', jbossCLITemplateName => 'snowflake', connectionOrResourceAdapterProperties => 'host=<your_account_name.snowflakecomputing.com,db=<database>,user-name=<user_name>,password=<user_password>,warehouse=<warehouse>') ;;

CALL SYSADMIN.createDataSource(name => 'snowflake', translator => 'snowflake', modelProperties => 'importer.defaultSchema=PUBLIC,importer.schemaPattern=PUBLIC,importer.useFullSchemaName=FALSE,importer.tableTypes="TABLE,VIEW"', translatorProperties => 'uploadMode=blob_azure,supportsNativeQueries=true,storageAccountName=<storageAccountName>,storageAccountKey=<storageAccountKey>,sasToken="<sasToken>",defaultEndpointsProtocol=https,tempFolder=d:\tmp\azure\,keepTempFiles=false') ;;

Connecting via Snowflake Storage Integration

To use the Snowflake Storage Integration, configure the following properties:

Property Description

externalStageName

Fully qualified name of the external stage predefined and configured in Snowflake

containerName

Name of the container defined in the Microsoft Azure Blob Storage account. Must be set to the same container as defined in externalStageName.

Please note that this property is case-sensitive as the container name is also case-sensitive in Microsoft Azure Blob Storage

Please keep in mind the following:

The value of the azureKey parameter should be the same as defined by the master_key parameter of the external stage;
By default, when azureKey is not set, the data is not encrypted.

Example

SQL

CALL SYSADMIN.createConnection(name => 'snowflake', jbossCLITemplateName => 'snowflake', connectionOrResourceAdapterProperties => 'host=<your_account_name.snowflakecomputing.com,db=<database>,user-name=<user_name>,password=<user_password>,warehouse=<warehouse>') ;;

CALL SYSADMIN.createDataSource(name => 'snowflake', translator => 'snowflake', modelProperties => 'importer.defaultSchema=PUBLIC,importer.schemaPattern=PUBLIC,importer.useFullSchemaName=FALSE,importer.tableTypes="TABLE,VIEW"', translatorProperties => 'uploadMode=blob_azure,supportsNativeQueries=true,storageAccountName=<storageAccountName>,storageAccountKey=<storageAccountKey>,externalStageName=<fully_qualified_name>,containerName=<case_sensitive_test_containrer_name>,defaultEndpointsProtocol=https,tempFolder=d:\tmp\azure\,keepTempFiles=false') ;;

Session Keepalive

By default, Snowflake terminates open sessions after four hours of inactivity. To prevent this, you can set the JDBC parameter CLIENT_SESSION_KEEP_ALIVE to TRUE:

SQL

CALL SYSADMIN.createConnection(name => 'dwh', jbossCLITemplateName => 'snowflake', connectionOrResourceAdapterProperties => 'host=<your_account_name.snowflakecomputing.com,db=<database>,user-name=<user_name>,password=<user_password>,warehouse=<warehouse>,jdbc-properties="CLIENT_SESSION_KEEP_ALIVE=TRUE"') ;;

CALL SYSADMIN.createDataSource(name => 'dwh', translator => 'snowflake', modelProperties => 'importer.defaultSchema=PUBLIC,importer.schemaPattern=PUBLIC,importer.useFullSchemaName=FALSE,importer.tableTypes="TABLE,VIEW"', translatorProperties => 'supportsNativeQueries=true,uploadMode=s3Load,region=<region>,bucketName=<bucket_name>,keyId=<key_ID>,secretKey="<secret_key>"') ;;
 
CALL SYSADMIN.setCurrentDWH('dwh', 'PUBLIC') ;;

Miscellanea

Indexes and native statistics are not supported by Snowflake data sources.