Configure a connector with the setup wizard¶

The Openflow connector setup wizard provides a guided, step-by-step experience for configuring gen 2 connectors. The wizard validates your inputs at each step so you can identify and fix configuration issues before completing the setup.

This topic applies to gen 2 connectors only. For gen 1 catalog connectors, see Openflow connectors. For an overview of gen 1 vs gen 2, see Openflow gen 1 and gen 2.

Supported connectors¶

The following gen 2 connectors are available in the catalog:

PostgreSQL CDC
MySQL and MariaDB CDC

Prerequisites¶

Before you start the setup wizard, make sure the following requirements are met:

You have a gen 2 Openflow deployment (BYOC or Openflow - Snowflake Deployment) with at least one gen 2 runtime. The wizard configures gen 2 connectors, which run on gen 2 runtimes only. To create deployment and runtime resources, see Quickstart: gen 2 Openflow. For BYOC cloud installation, Snowflake deployment networking, deployment setup, and runtime parameters, see Set up Openflow - BYOC, Set up Openflow - Snowflake Deployment - Task overview, or Set up Openflow - Snowflake Deployment: Create runtime as applicable. If you are unsure whether your resources are gen 1 or gen 2, see Openflow gen 1 and gen 2.
For Openflow - Snowflake Deployments, your PostgreSQL database must be reachable through an external access integration (EAI) that references a network rule permitting egress to your PostgreSQL hostname and port (the connector uses a customer-specific endpoint; see PostgreSQL). Grant USAGE on the EAI to the runtime’s EXECUTE_AS_ROLE (see Create an execute-as role). For setup steps, see Configure external access in the gen 2 quickstart or Set up Openflow - Snowflake Deployment: Configure allowed domains for Openflow connectors. If you use Openflow BYOC, configure outbound connectivity in your cloud environment so the runtime can reach PostgreSQL; Snowflake EAIs are not used in that deployment model.
Any required secrets (for example, database passwords or API keys) are created as Snowflake secrets and accessible to the runtime’s EXECUTE_AS_ROLE (see Create an execute-as role).
You have CREATE OPENFLOW CONNECTOR on the schema that the runtime is in, USAGE on the runtime, and USAGE on the runtime’s database and schema.
You have the required permissions for the connector you want to install, and have completed any connector-specific source database prerequisites: PostgreSQL CDC | MySQL and MariaDB CDC. For Snowflake account setup (destination database, warehouse, and secrets), follow the Snowflake account setup on those pages.

Create and configure a connector¶

In the navigation menu, select Ingestion » Openflow.
Select Launch Openflow.
In the Featured Connectors section, select View More Connectors.
Find a gen 2 connector in the catalog.
Select a runtime to install the connector to. The list is filtered based on the runtime size requirements needed to run the connector.
Select Install.
Complete the wizard steps. The specific steps vary depending on the connector you selected. Documentation for every step is present within the wizard.
Upon completion, select Create Connector.
You’re redirected to the Installed Connectors page. Once the connector has finished installing, select Start from the menu for your connector to start moving data.

At any point during the wizard, you can select Save and Close to preserve your configuration changes without applying them to the connector. To apply your saved changes later, select Edit from the menu for that connector on the Installed Connectors tab, step through the wizard to the end, and select Apply.

To create and configure the same connector types with SQL instead of the wizard, see Configure a gen 2 connector with SQL.

Edit an existing connector¶

To edit a connector that you’ve already created, go to the Installed Connectors tab and select Edit from the menu for that connector. This reopens the setup wizard with your existing configuration.

Input validation¶

The wizard provides a button that allows you to validate your input at each step. Validation checks include:

Connection tests: The wizard attempts to connect to your data source using the provided credentials and displays the result.
Permission checks: The wizard verifies that the specified Snowflake user has the required privileges on the target database, schema, and table.
Format validation: The wizard checks that values such as hostnames, ports, and database names are in the expected format.

If a validation check fails, the wizard highlights the field and displays a message explaining the issue. You can move to the next step even when intermediate validation fails; only the final validation before Create Connector (or Apply) must pass.

Troubleshooting¶

When you select Verify and validation fails, the wizard displays troubleshooting information directly in the UI explaining what went wrong and how to fix it. The panel on the right side of the wizard also contains detailed information about what is required for each property.

Verify ingestion (dashboard, destination database, read-only canvas)¶

After you Start the connector, confirm that data is flowing:

Use the Openflow Connectors Dashboard (Ingestion » Openflow » Connector Observability) for health, throughput, and errors.
In Snowflake, query the destination database (the tables or schemas the connector writes to) to verify that new rows or changes are arriving as you expect.
Open the runtime canvas to inspect processors, queues, FlowFiles, and bulletins for your connector’s process group—a read-only operational view. From the dashboard, select Go to canvas for that connector; or open the canvas from Runtimes in Openflow. You cannot configure the connector on the canvas for setup wizard connector types; use Installed Connectors » Edit or the connector API for any configuration changes.

For starting and stopping ingestion, removal actions, and how the canvas compares to catalog-installed connectors, see Manage the gen 2 Openflow connector lifecycle.

Next steps¶

After you configure a connector with the wizard, you can: