@kapa.ai what's the best way to manage connector versions (source/destination)? Is it using the API? Terraform? UI is not acceptable as I need some degree of automation

Both the Terraform provider and the API are solid options for automating connector version management in Airbyte. Here's a breakdown to help you choose: Terraform Provider The Airbyte Terraform provider is built on top of the Airbyte API and is the recommended approach if you're already managing infrastructure as code. You can pin a connector to a specific version using the

airbyte_connector_configuration

data source: [Terraform Registry]

data "airbyte_connector_configuration" "github" {
  connector_name    = "source-github"
  connector_version = "2.0.0"

  configuration = {
    repositories = ["airbytehq/airbyte"]
    credentials = {
      option_title          = "PAT Credentials"
      personal_access_token = var.github_pat
    }
  }
}

resource "airbyte_source" "postgres" {
  name          = "Postgres"
  definition_id = data.airbyte_connector_configuration.postgres.definition_id
  workspace_id  = var.workspace_id
  configuration = data.airbyte_connector_configuration.postgres.configuration_json
}

When

connector_version

is omitted, the latest version is used automatically. This gives you explicit version control directly in your IaC configuration. [Terraform Registry] Airbyte API The API is the underlying layer that powers the Terraform provider. As of Airbyte 1.6+, the API also supports creating and updating custom connectors (YAML or Docker-based) via dedicated endpoints, e.g.: [release notes] •

POST /workspaces/{workspaceId}/definitions/declarative_sources/

— create a YAML source definition •

PUT /workspaces/{workspaceId}/definitions/declarative_sources/{definitionId}

— update it Recommendation • Use Terraform if you want version-controlled, declarative management of your sources and destinations with explicit version pinning. It's the most automation-friendly approach for infrastructure teams. [Terraform docs] • Use the API directly if you need more programmatic flexibility (e.g., scripting, CI/CD pipelines) or if you're managing custom connectors across workspaces. [release notes] Both approaches support automation well; Terraform adds the benefit of state management and drift detection on top of the API. 👍 0 button 👎 0 button ℹ️ Tip: Mention @kapa.ai in the thread for follow-ups or DM the app for 1-to-1 help.

@kapa.ai how to import existing destinations using terraform? How to get any needed IDs?

Importing Existing Destinations into Terraform The Airbyte Terraform provider supports importing existing destinations into Terraform state. There are two methods: Method 1:

import

block (Terraform v1.5.0+)

import {
  to = airbyte_destination.my_airbyte_destination
  id = "<destination-id>"
}

[airbyte_destination import] Method 2:

terraform import

CLI command

terraform import airbyte_destination.my_airbyte_destination "<destination-id>"

[airbyte_destination import] Both methods work for the generic

airbyte_destination

resource as well as typed resources like

airbyte_destination_postgres

,

airbyte_destination_s3

,

airbyte_destination_bigquery

, etc. [destination_postgres import; destination_s3 import] ------------------------- Getting the Destination ID The knowledge sources don't explicitly describe how to retrieve destination IDs, but the standard approaches are: 1. Airbyte UI: Navigate to the destination's settings page — the destination ID is typically visible in the browser URL (e.g., `.../destinations/`destination-id). 2. Airbyte API: Use the

GET /destinations

endpoint to list all destinations and find the relevant ID. ------------------------- Note on

airbyte_destination_definition

If you're importing a destination definition (rather than a destination connector instance), the import syntax is slightly different — it requires both an

id

and a

workspace_id

as a JSON-encoded string:

terraform import airbyte_destination_definition.my_airbyte_destination_definition '{"id": "...", "workspace_id": "..."}'

[destination_definition import] 👍 0 button 👎 0 button ℹ️ Tip: Mention @kapa.ai in the thread for follow-ups or DM the app for 1-to-1 help.