Tutorial: Managing Connection using Fivetran REST API

This guide explains how to manage a single connection through its entire lifecycle using the Fivetran REST API. It covers creation, connectivity checks, optional certificate and SSH fingerprint approval, schema management, scheduling and triggering syncs, ongoing monitoring, and eventual deletion.

How to create a connection

Use case: Create a connection to your data source: a SaaS application, database, file, or event-capturing service.

• Call the Create a Connection endpoint with service, group_id, optional paused, and any connection-specific config. On this endpoint's documentation page, you can choose the connector type (service in the request payload) and the language of your choice to review the request payload as well as the response. The response returns the connection id, an array of setup_tests, and default scheduling info.

How to verify connectivity and re-run tests

Use case: After rotating credentials or changing firewall rules, re-run tests to confirm the connection is still healthy.

• Inspect the initial setup_tests to confirm host and credential checks.
• Call the Run Connection Setup Tests endpoint whenever you modify network or credentials to ensure the connection still works.

How to approve certificates or fingerprints

Use case: Necessary when connecting to databases with self-signed certificates or verifying the identity of an SSH host.

• Trust a TLS certificate by calling the Approve a Certificate for the Connection endpoint. For more details on TLS certificates, see the Certificate Verification documentation section.
• Approve SSH fingerprints by calling the Approve a Fingerprint for the Connection endpoint using the failing test’s hash and public_key.

How to manage connection schema before syncing

Use case: Only sync required tables, avoiding PII or large audit logs, and reduce network traffic.

• Reload schema configuration by calling the Reload a Connection Schema Config endpoint.
• Set "exclude_mode": "EXCLUDE" so nothing syncs until explicitly enabled.
• Enable or disable tables, or columns by calling the Update a Connection Schema Config endpoint for granular control.
• (For database connectors with multiple schemas) Enable or disable schemas by calling the Update a Connection Database Schema Config endpoint.

How to enable and schedule syncs

Use case: Adjust sync frequency to balance freshness with API costs, or run a historical reload when initial data was incomplete.

• Unpause and configure sync schedule by calling the Update a Connection endpoint:
  • paused — start or stop the connection
  • sync_frequency - Set the sync frequency interval in minutes. For example, 1440 runs a sync every 24 hours.
  • daily_sync_time - Set the daily sync start time. For example, 14:00. Applicable only when sync_frequency is 1440.
  • schedule_type - Set the sync scheduling type. Specifying auto means syncs run at the defined frequency_sync, and manual means you trigger your syncs on demand.
  • is_historical_sync - Set to true to make the next sync a historical re-sync. The value resets to false after the run, and next syncs are incremental.
  • name - Rename the connection.
  • group_id - Move data to a different destination group.
  • config - Update credentials or connector-specific settings.
• Trigger syncs on demand:
  • Run an incremental sync by calling the Sync Connection Data endpoint.
  • Run a historical re-sync by calling the Re-sync Connection Data (Historical Sync) endpoint.

How to maintain connection

Use case: Maintain connections by updating its parameters and re-running setup tests.

• Update credentials, schedule, or other parameters through repeated calls to the Update a Connection endpoint.
• Re-run setup tests by calling the Run Connection Setup Tests endpoint) after infrastructure changes.

How to remove connection

Use case: Remove connection for deprecated data sources.

Remove the connection permanently by calling the Delete a Connection endpoint when it’s no longer required.