Google Cloud Function that enables stated APIs for newly created projects. GCP API Enabler saves your time, especially when you need to create multiple GCP projects at once.
GCP API Enabler has two handler methods - one that gets triggered with simple GET request, and another one that can be triggered from Google Cloud Pub/Sub topic when new project in organization is created. In the documentation below, we will show you how to setup a function that will be triggered when a new project is created under your Google Cloud organization.
The one that gets triggered with GET request will enable stated APIs for all your organization project, and because of that it can take a while. It can also accept project_number parameter and it that case it will enable APIs only for a specific project. This function is commented out in serverless.yml and it will not be deployed by default (because of security reasons), but it can be useful for testing.
The other one, triggered by the Pub/Sub topic when new project is created, will enable APIs only for that project.
You can also trigger your cloud function directly from "Testing" tab, by sending data as:
{
"project_number": "testing-project-123"
}
If you do not have it already, install gcloud SDK.
git clone git@github.com:kiwicom/gcp-api-enabler.git
cd gcp-api-enabler
pip install -r requirements.txt
npm installYou will deploy GCP API Enabler to Google Cloud Platform using Serverless framework. But first, make sure you configure everything mentioned below. Most of the actions can be fulfilled using gcloud SDK, but for some, you will need to handle it manually.
If you have access to your organization and you do not have Billing Account yet, you will need to create a new Billing Account first. Otherwise, skip this step.
If you do not have an existing GCP project where you will setup API enabler, you will need to create a new one, either manually or with gcloud SDK.
gcloud projects create your-project-id --name="Your project name"Otherwise, skip this step.
In order for Serverless framework to work, you will need to enable this necessary APIs.
Additionally, in order for this Cloud Function to work, enable this necessary APIs:
gcloud services enable cloudfunctions.googleapis.com
gcloud services enable deploymentmanager.googleapis.com
gcloud services enable storage-component.googleapis.com
gcloud services enable logging.googleapis.com
gcloud services enable cloudresourcemanager.googleapis.com
gcloud services enable serviceusage.googleapis.comIn order to create a new service account that you will use to deploy API Enabler, follow Serverless documentation.
If you already have sufficient service account, skip this step.
At runtime, Cloud Functions defaults to using the service account PROJECT_ID@appspot.gserviceaccount.com, which has the Editor role on the project.
Since API Enabler needs to access all the projects under the organization, be sure you grant editor role to this service account (PROJECT_ID@appspot.gserviceaccount.com) on the organization level. This is needed because API Enabler needs to get the list of services in other projects and needs to have permissions to enable them.
- Create Pub/Sub topic
gcloud pubsub topics create topic-name- Create aggregated export within your organization
gcloud logging sinks create sink-name \
pubsub.googleapis.com/projects/your-project-id/topics/topic-name --include-children \
--organization=YOUR_ORGANIZATION_ID --log-filter="resource.type=project AND protoPayload.methodName=CreateProject"Output of this method will be something like:
Please remember to grant 'serviceAccount:o123456789012-123456@gcp-sa-logging.iam.gserviceaccount.com' Pub/Sub Publisher role to the topic.
- Manually add service account from Step 2 as a publisher role to the Pub/Sub topic created in Step 1.
- Try to test your topic to see if it works correctly.
cp .env.example .envIn .env, adjust PROJECT_ID, GCLOUD_CREDENTIALS, TOPIC_NAME and SERVICES_TO_ENABLE.
PROJECT_ID- your (new) project id.GCLOUD_CREDENTIALS- path to gcloud credentials downloaded above.TOPIC_NAME- Pub/Sub topic name you created above.SERVICES_TO_ENABLE- list of services you would like to enable by default, separated by comma(,).
SERVICES_TO_ENABLE will be deployed as an environmental variable to your Google Cloud Function. If you are changing a list of services, you will need to redeploy your code.
serverless deploy -vBug reports and fixes are always welcome!
Tests are run with pytest. Install into virtual environment
requirements.txt and test-requirements.txt and run in shell command pytest
Code is formatted by Black.