endpointsapis/

Service Infrastructure Codelab

This is an example API that demonstrates how to design a networked API following API Design Guide, implement the API as an App Engine app, and integrate the API with Google Cloud's Service Infrastructure. It is intended for service providers who want to build services using Google Cloud.

NOTE: This API is also used for Service Infrastructure reference documentation.

Glossary

For terminology used in this example, see API Glossary.

Installation

You need to install gcloud, go, protoc, and oauth2l tools on your local machine in order to build this example.

Run the following commands to install additional Go packages that will be used in this example API:

go get golang.org/x/oauth2/google
go get google.golang.org/api/servicecontrol/v2

Introduction

This API uses a proto file to define the API surface, and a YAML file to configure the API service. The API backend is built as an App Engine app written in Go, and uses the Service Control API for deeper integration with Service Infrastructure and Google Cloud.

This example API uses project id endpointsapis and service name endpointsapis.appspot.com. The name endpointsapis has no special meaning. If you want to build this example API by yourself, you must use your own project id and service name. For how to create a project, see Getting Started.

WARNING: Some features used in this example are private, and you need to join private preview programs to use such features.

Design

This example API design follows Google's API Design Guide. It uses a proto file, ./v1/workspace.proto, to define the gRPC API surface, and uses google.api.http annotation to define the REST API mapping.

For more information, see the proto definition and API Design Guide.

Configuration

To build a networked API and expose it to consumers, you need to configure it as an API service. The configuration includes the API title, the network addresses, the authentication requirements, and many other aspects.

If you create an API service using Google Cloud's Service Infrastructure, you need to create a servie configuration using the google.api.Service schema. The service configuration is the yaml representation of google.api.Service, see ./endpointsapis.yaml.

type: google.api.Service  # The schema to use; always "google.api.Service".
name: endpointsapis.appspot.com
title: Endpoints APIs
producer_project_id: endpointsapis
apis:                     # The supported API interfaces.
- name: google.example.endpointsapis.v1.Workspaces

After you have created your API definition and service configuration, you need to push them to the Service Management API, which will distribute the information to Google Cloud infrastructure backends.

Run the following commands to generate proto descriptor and push the descriptor and service configuration to the Service Management API.

# Generate proto descriptors from the proto files.
protoc --include_source_info --include_imports --descriptor_set_out=service.descriptors v1/workspace.proto

# Push the proto descriptors and yaml files to the Service Management API.
gcloud endpoints services deploy service.descriptors endpointsapis.yaml

For more information, see Manage Service Configurations.

Implementation

Create an App Engine app that implements your API and deploy the app, see Building a Go App on App Engine. The implementation of this code is in goapp/main.go. Run the following command to deploy the code:

gcloud app deploy

Integration

The Service Control API is the frontend of System Infrastructure control plane. It provides admission control and telemetry reporting functionality to any service that is integrated with Service Infrastructure, specifically:

• For each request, the service calls the Check method for admission control.
• For each response, the service calls the Report method for telemetry reporting.

For more information, see Admission Control and Telemetry Reporting.

In this example API, the app calls the Service Control API v2 for each request and response, see goapp/main.go. You need to make sure the App Engine service account has the necessary permissions to run the service. For more information, see Access Control.

NOTE: The telemetry reporting feature is not yet available as of July 2020.

Verification

Follow Getting Started to set up a client project for calling your API service. If you want to run your test locally, you need to create an application credential in your client project and download it to your local machine. The credential can be either an OAuth client or a service account key in JSON format.

Use oauth2l to send a test request to your App Engine app using your application credentials. oauth2l supports different credential types in different environments.

curl -H "$(oauth2l header --credentials creds.json --scope cloud-platform,userinfo.email)" https://endpointsapis.uc.r.appspot.com/v1/projects/1020789478714/locations/1/workspaces

You should see the admission control results in the HTTP response and the telemetry reporting results in Google Cloud Console. The producer metrics are shown on the Endpoints page. The consumer metrics are shown on the APIs & Services page.

Summary

This example shows how to build a simple API service using Google Cloud infrastructure. If you have any suggestions, please send us pull requests directly.