Skip to main content

Snowflake Destination Connector

The Snowflake destination connector writes extracted and transformed data from DataSync into a Snowflake database. Snowflake is a fully managed, cloud-based OLAP database built for batch processing and large datasets. DataSync supports key pair authentication for Snowflake, using a public and private key pair instead of a password.

Because of how Snowflake is built, a few standard database features work differently here. Snowflake does not support triggers or indexes, and it only enforces NULL and NOT NULL constraints on standard tables. This means that when you add a Snowflake source table to an extraction, any indexes from the original source are not carried over, and you cannot create indexes manually afterward.

Once your destination connection is ready, create your extraction to define what data gets loaded and how.

How Snowflake stores and processes your data

  • Storage and compute are separate and scale independently of each other.
  • Data is automatically partitioned into chunks of 50 to 100 MB, stored in a columnar format. Each partition tracks minimum and maximum values, letting Snowflake skip irrelevant partitions during a query.
  • Queries run in parallel across multiple compute nodes instead of on a single machine.

Create the destination connection in DataSync

  1. Log in to DataSync.
  2. From the welcome screen, select Connections.
  3. Next to Destination Connections, click New.
  4. Select Snowflake.
  5. Enter all required connection properties.
  6. (Optional) In the Additional Connection Properties panel, click Add property and enter any advanced parameters.
  7. Click Save.

Connection properties

PropertyWhat to enter
DescriptionUnique name for the connection. Example: Snowflake.
AccountFull account identifier. In Snowflake, open the account selector, select your account, then select View Account Details. The identifier appears under Account identifiers. Example: xy12345.us-east-1.
DatabaseName of the database to write data to. Example: AnalyticsDB.
Authentication ModeAuthentication method for the connection.
  • Basic for username and password authentication.
  • KeyPair for key-pair authentication. Snowflake requires at least a 2048-bit RSA key pair.
UsernameBasic only Snowflake database username.
PasswordBasic only Password for this account. The user type must be LEGACY_SERVICE. Snowflake is phasing out single-factor Basic authentication, so use this method only if no other option is available.
Key UserKeyPair only Snowflake database username associated with the public key.
Private Key FileKeyPair only Path to the private key file associated with the public key.
PassphraseKeyPair only Password protecting the private key file. Leave blank if the key is not encrypted.
TimeoutTime in seconds to wait before a connection attempt or query execution times out.
Insecure ModeDisables certificate revocation list checks. Use only as a temporary workaround during OCSP-related service disruptions.
Disable RetryDisables automatic retries on failed HTTP requests to Snowflake API endpoints.
Max HTTP RetriesMaximum number of retry attempts for failed HTTP requests when retries are enabled. Range: 1 to 999. Default: 7.
Retry TimeoutTotal time in seconds allowed for requests covered by the retry policy. Default: 300.

Additional connection properties

Use this panel to enter connection string properties not available in the Connection Properties panel. For sensitive values such as passwords, set the type to Encrypted. The value is hidden in the interface and stored encrypted in the back end.

PropertyWhat to enter
PropertyConnection string property that defines the action or behavior. Example: ReadOnly.
ValueValue for the property. Example: True.
TypeVisibility settings: Visible or Encrypted.

Key pair authentication example setup

Snowflake key pair authentication destination connection in DataSync
Completed Snowflake destination connection in DataSync using key pair authentication, with all properties and settings filled in.