terrabridge

Terrabridge bridges the gap between Terraform and applicatoin code. With Terrabridge you simply provide your terraform state file to the library and all information will be loaded allowing you to easily access your resources. This allows you to truly keep your infrastructure configuration in one place.

Once your terraform and application code are in sync Terrabridge makes it dead simple to connect to your resource by providing easy ready to use clients for your resources such as SQLAlchemy for relational databases, or reading secret values from secret storage.

from terrabridge.gcp import SecretManagerSecret

sec = SecretManagerSecret("secret", state_file="terraform.tfstate")
print(sec.version().decode("utf-8"))

Installation

pip install terrabridge

To use the GCP clients you will need to install the GCP extras:

pip install terrabridge[gcp]

To use the AWS clients you will need to install the AWS extras:

pip install terrabridge[aws]

Usage

Basic Usage

Terrabridge can be used by providing your state file to the library. The state file can be local, stored in an S3 bucket, or in a GCS bucket. Terrabridge will then parse the state file into a python object that can be consumed by your application code.

For example if you had the below terraform code that creates and manages a GCS bucket you can easily access the bucket from your application code using the GCSBucket class. All terrabridge objects take in a state file and the name of the resource you assigned in terraform. All attributes that are available in terraform are now available on your terrabridge object.

from terrabridge.gcp import GCSBucket

bucket = GCSBucket("bucket", state_file="terraform.tfstate")
# Fetches the remote bucket.
print(bucket.url)
print(bucket.id)
print(bucket.name)
bucket = bucket.bucket()
print(bucket.get_iam_policy().bindings)

resource "google_storage_bucket" "bucket" {
    name = var.bucket_name
    location = "US"
}

Global State File

If all of your terrabridge objects use the same state file you can set a global state file, and avoid passing it to each object:

import terrabridge
from terrabridge.gcp import GCSBucket

terrabridge.state_file="terraform.tfstate"

bucket = GCSBucket("bucket")

Remote State File

If your state file is stored in an S3 bucket or GCS bucket you can pass the bucket name and key to the state_file argument. Terrabridge will then download the state file and parse it.

from terrabridge.gcp import GCSBucket
from terrabridge.aws import S3Bucket

gcs_bucket = GCSBucket("bucket", state_file="gs://my-bucket/terraform.tfstate")
s3_bucket = S3Bucket("bucket", state_file="s3://my-bucket/terraform.tfstate")

Examples

S3 Bucket

Easily connect and read data from a S3 bucket, that is defined in terraform.

TODO

GCS Bucket

Easily connect and read data from a GCS bucket, that is defined in terraform.

Python:

from terrabridge.gcp import GCSBucket

bucket = GCSBucket("bucket", state_file="terraform.tfstate")
bucket = bucket.bucket()
print(bucket.get_iam_policy().bindings)

Terraform:

variable "gcp_project_id" {
    type = string
    description = "The GCP project to deploy resources into."
}

variable "bucket_name" {
    type = string
    description = "Name of the bucket."
}


provider "google" {
    project = var.gcp_project_id
    region  = "us-central1"
    zone    = "us-central1-a"
}

resource "google_storage_bucket" "bucket" {
    name = var.bucket_name
    location = "US"
}

Cloud SQL Postgres Database

Use SQLAlchemy to connect to a managed Cloud SQL Postgres database with one function call.

Python:

import datetime
import uuid

from sqlalchemy import Column, DateTime, Integer, select
from sqlalchemy.ext.asyncio import AsyncAttrs
from sqlalchemy.orm import DeclarativeBase

from terrabridge.gcp import CloudSQLDatabase, CloudSQLUser


class Base(AsyncAttrs, DeclarativeBase):
    pass


class StorageUser(Base):
    __tablename__ = "users"

    # Autopoluated fields
    id = Column(Integer, primary_key=True, default=uuid.uuid4)
    created_at = Column(DateTime, default=datetime.datetime.utcnow)


db = CloudSQLDatabase("postgres_database", state_file="terraform.tfstate")
user = CloudSQLUser("postgres_user", state_file="terraform.tfstate")
engine = db.sqlalchemy_engine(user)

Base.metadata.create_all(engine)

conn = engine.connect()
print(conn.execute(select(StorageUser)).all())

Terraform:

variable "gcp_project_id" {
    type = string
    description = "The GCP project to deploy resources into."
}

provider "google" {
    project = var.gcp_project_id
    region  = "us-central1"
    zone    = "us-central1-a"
}

resource "google_sql_database_instance" "postgres_sql_instance" {
    name = "terrabridge-testing-instance-mysql"
    project = var.gcp_project_id
    database_version = "POSTGRES_15"
    region = "us-central1"
    settings {
        tier = "db-custom-1-3840"
    }
}

resource "google_sql_database" "postgres_database" {
    name = "terrabridge-testing-database"
    project = var.gcp_project_id
    instance = google_sql_database_instance.postgres_sql_instance.name
}

resource "google_sql_user" "postgres_user" {
    name = "terrabridge-testing-user"
    project = var.gcp_project_id
    instance = google_sql_database_instance.postgres_sql_instance.name
    password = "terrabridge-testing-password"
}

Supported Providers and Languages

Python in the first language we support however we plan to support more languages in the future. We are always happy to accept contributions for new languages and providers.

	python	golang	java	typescript
gcp	✅	❌	❌	❌
aws	🚧	❌	❌	❌
azure	❌	❌	❌	❌

GCP Supported Resources

TODO: add links to docs

• BigQuery Dataset
• BigQuery Instance
• BigTable Instance
• BigTable Table
• Cloud SQL Database Instance
• Cloud SQL Database
• Cloud SQL User
• Cloud Tasks Queue
• GCS Bucket
• Pub/Sub Lite Topic
• Pub/Sub Lite Subscription
• Pub/Sub Topic
• Pub/Sub Subscription
• Secret Manager Secret

AWS Supported Resources

TODO