Skip to main content
A Snowflake connection lets Prophecy read from and write to tables in your Snowflake account. Use it with Source and Target gems, browse tables in the Environment browser, and process Snowflake data in Prophecy pipelines. This page covers Snowflake connections used to access data. To use Snowflake as the compute engine for a fabric, see Create a Snowflake fabric.

Connection types

You can use Snowflake in two ways in Prophecy.

Feature support

The following table shows which Prophecy features each connection type supports.

Prerequisites

Prophecy runs all data operations, such as reads and writes, with the Snowflake credentials you provide in the connection. Grant your Snowflake user or role the privileges for the operations your pipelines perform. When you write data through a Snowflake connection, Prophecy uploads Parquet files to your user stage and then loads them into Snowflake tables. This is why write operations require file format and stage access.

Connection parameters

Starting October 2026, Snowflake enforces the deprecation of password-based sign-in for service users. Connections that authenticate with a username and password stop working once Snowflake enforces the cutover. Prophecy flags affected connections with a Password auth badge on the Connections tab. Move these connections to key-pair, OAuth, or OIDC authentication.
To create a connection with Snowflake, enter the following parameters:

Key-pair authentication

Key-pair authentication signs in to Snowflake with an RSA private key that you store in a Prophecy secret. Before you begin, assign the public key to the Snowflake user that the connection uses. The private key must be an unencrypted PKCS#8 key. If your key is encrypted, convert it with OpenSSL. To configure key-pair authentication:
  1. In the fabric, create a Text secret. Click Or Upload File and upload the private key file. Upload the file instead of pasting the key into Value. Pasting can remove the line breaks the key requires.
  2. In the connection, under Authentication method, select Key Pair.
  3. Enter the Username of the Snowflake user that has the public key assigned.
  4. For Private Key, select the secret you created.
  5. Leave Private Key Passphrase (Optional) empty.
Key-pair authentication doesn’t support passphrase-protected private keys. Use an unencrypted private key.
Anyone with the unencrypted key file can sign in as its Snowflake user. After you upload the key to the secret, delete the local file or store it securely.

OIDC authentication

OIDC authentication signs in to Snowflake as a service principal whose credentials are issued by your identity provider. Prophecy exchanges these credentials for a bearer token through an OIDC federator, caches the token, and reconnects with a new token before the old one expires. OIDC doesn’t require an app registration or an interactive sign-in. Unlike key-pair authentication, the certificate and private key come from your identity provider, not from a key pair registered on a Snowflake user. The connection doesn’t take a username. Instead, Snowflake maps a claim in the token to a Snowflake user through an External OAuth security integration, which must exist before you create the connection. To configure OIDC authentication:
  1. In the fabric, create secrets for the service principal’s certificate and private key, and for its client secret if you have one.
  2. In the connection, under Authentication method, select OIDC.
  3. Enter the Auth Server URL of the token issuer.
  4. Enter the Resource that the token is requested for.
  5. Enter the Service Principal Client ID.
  6. Optionally, for Service Principal Client Secret, select the client secret.
  7. For Certificate, select the certificate secret.
  8. For Private Key, select the private key secret.
The certificate and private key are required because the federator signs a client assertion with the private key. A client secret alone cannot authenticate the connection.
OIDC authentication requires an OIDC federator URL in the Prophecy deployment configuration. If the federator URL isn’t configured, connections that use OIDC fail with the error OIDC federator URL is not configured. Contact your Prophecy administrator to configure it.

Data type mapping

When an Ingress/Egress connection reads Snowflake data into pipelines that run on another SQL warehouse, such as Databricks or BigQuery, Prophecy converts Snowflake-specific data types to compatible types. The following table shows how Snowflake data types map to Databricks and BigQuery types.
Learn more in Supported data types.

Limitations

Snowflake connections have the following data type limitations:
  • Prophecy reads OBJECT, ARRAY, and VARIANT columns as STRING.
  • Prophecy does not support writing BINARY columns.
  • On Databricks and BigQuery, Prophecy maps DECIMAL and NUMERIC to floating-point types, which can reduce precision. Avoid this conversion for values that require exact precision, such as currency.
  • On Databricks and BigQuery, Prophecy maps NUMBER to an integer type, so values with a decimal scale lose their fractional digits.
  • On Databricks, Prophecy maps TIME to STRING.

Sharing connections within teams

Prophecy stores connections in fabrics, which are assigned to specific teams. When you add a Snowflake connection to a fabric, every team member with access to that fabric can use the connection in their projects. Team members don’t authenticate separately. They inherit the access and permissions of the stored connection credentials.
Anyone on the team has the same permissions as the stored credentials, including access to sensitive data if those credentials allow it.To limit access to high-sensitivity connections, create a dedicated fabric and team for them so only approved users can use those credentials.