Connections and connectors

A connection is a reusable, permissioned reference to an external system. It keeps endpoint and namespace settings together while sensitive values remain in Vegalake Secrets.

Connector availability

The connector catalog shows enabled connectors, supported sync modes and whether a connector can be a source, destination or both. PostgreSQL is currently enabled for production QuickFlows. GitHub source and Airtable destination definitions remain preview until your workspace shows them as enabled.

Connector	Category	Supported use	Availability
PostgreSQL	Databases	Source or destination	Enabled
GitHub	APIs and applications	Source schema	Preview
Airtable	APIs and applications	Destination schema	Preview

Browse connector guides by type for database, API/application and filesystem or object-storage documentation coverage.

Trust the workspace catalog

Connector availability changes independently from this documentation. VegaFlow → Connectors is the authoritative list for your organization and region.

Create a connection

Choose a connector

Open VegaFlow → Connections → New connection and choose an enabled source or destination.

Add configuration and secrets

Enter a unique, descriptive name, optional description and the fields required by the connector schema. Select existing secret references or create a vault and secret inline when permitted.

Choose network access

Select the VegaFlow compute environment and private network route that will reach the system.

Test and save

Validate DNS, TLS, authentication and the minimum required permissions before saving the reusable connection.

PostgreSQL

For a PostgreSQL source or destination, configure:

Host / port       db.example.internal:5432
Database          orders
User              vegalake_sync
Password          Secret: orders-sync-password
TLS mode          verify-full
Allowed schemas   public, billing

Source users need connection, schema usage and SELECT privileges. Destination users need schema and table write privileges appropriate to the selected write mode. Select only sync modes advertised by the discovered stream.

The PostgreSQL connector guide covers the exact database fields, SSL modes, password authentication and SSH tunnels.

Keep per-flow behavior out of the connection. Table selection, cursor columns, primary keys, destination mapping, schema evolution and write mode belong to the QuickFlow stream configuration.

Discover data

Open the connection and choose Discover with a running compute environment. VegaFlow returns namespaces, streams, fields, types and supported sync modes without creating a flow. Narrow the allowed namespaces for very large catalogs.

Discovery is repeated when a QuickFlow run prepares, so newly added or removed source streams can be reconciled safely. A retry of failed streams refreshes discovery but does not automatically add a new stream to that retry.

API

GET  /api/v1/vegaflow/orgs/{org_id}/workspaces/{workspace_id}/vegaflow/connectors
GET  .../connectors/{connector_id}/properties
POST .../connections/validate
POST .../connections
GET  .../connections/{connection_id}/quickflow/catalog?cluster_id={cluster_id}

Read the connector’s properties schema before building a form or deployment payload. Connector fields and supported authentication options are versioned.

Create and update requests require a name, connector ID, connector type and configuration object. Disabled connectors and ID/type mismatches are rejected. Check organization/workspace scope, name conflicts and secret permissions when validation fails.

Updates and deletion

Updating a connection creates a new configuration version. Revalidate after changing network, database, connector or credential settings. Existing run history retains the configuration identity it used. Deletion is blocked while an active QuickFlow references the connection, then retained as an auditable soft deletion.

Connections and connectors

On this page