Firebase

Firebase Source Connector

Description

The Firebase Source Connector allows reading data from Google Firebase Realtime Database via its REST API. It supports extracting node structures into SeaTunnel internal rows using defined schema fields.

Key Features

• Batch
• Stream
• Exactly-Once
• Column Projection

---

Options

Name	Type	Required	Default	Description
url	String	Yes	-	The base URL of your Firebase Realtime Database (e.g., https://<DATABASE_NAME>.firebaseio.com).
path	String	Yes	-	The JSON node path to read from (e.g., users or logs/2026).
service_account_path	String	No	-	Absolute path to the Google Service Account JSON key file for OAuth2 service authentication.
credentials	String	No	-	Base64-encoded Service Account JSON credentials content.
database_secret	String	No	-	Legacy Firebase database secret key or Web API token.
timeout_ms	Integer	No	10000	HTTP request timeout in milliseconds.
query_params	Map	No	-	Additional REST API query parameters to pass with requests.
schema	Config	Yes	-	The target table schema definition mapping JSON keys to SeaTunnel data types.
common-options	Config	No	-	Source common options. Refer to Source Common Options for details.

---

Data Type Mapping

The Firebase connector converts incoming JSON node structures into SeaTunnel internal data types:

Firebase / JSON Type	SeaTunnel Data Type
String	STRING
Number (Integer)	INT / BIGINT
Number (Floating point)	FLOAT / DOUBLE
Boolean	BOOLEAN
Object (Nested node)	ROW / MAP
Array	ARRAY

---

How to Set Up

1. Authentication Options

The connector supports three authentication methods:

• service_account_path: Point to a local service account JSON key file downloaded from Firebase Console > Project Settings > Service accounts.
• credentials: Provide the Base64-encoded content of a service account JSON file (useful for CI/CD or cloud secret injection).
• database_secret: Pass a legacy database secret key or Web API token.

2. Path & Record Structuring

• Collection Path (path = "users"): The connector fetches child node objects (e.g., user_101, user_102) and emits each child object as an individual row.
• Single Record Path (path = "users/user_101"): The connector automatically aggregates child primitive fields (e.g., name, role) into a single output row matching the schema.

---

Example

Reading a Collection Path

env {
  parallelism = 1
  job.mode = "BATCH"
}

source {
  Firebase {
    url = "https://my-app-default-rtdb.firebaseio.com"
    path = "users"
    service_account_path = "/etc/seatunnel/firebase-credentials.json"
    timeout_ms = 5000

    schema {
      fields {
        name = "string"
        role = "string"
      }
    }
    plugin_output = "firebase_users"
  }
}

sink {
  Console {
    plugin_input = "firebase_users"
  }
}

Changelog

Change Log

Change	Commit	Version
[Feature][Connector-V2] Add Firebase Source Connector		Unreleased