XTechnology - Technology is beautiful, let's discover it together!

How to develop, build, and deploy Node.js NestJS GRPC microservices with Pulumi and Google Cloud Platform

The workshop gives a practical perspective of key principles needed to develop, build, and maintain a set of microservices in the Node.js stack. It covers specifics of creating isolated TypeScript dockerized services using NestJS framework and a monorepo approach with turborepo. The workshop includes an overview and a live exercise to create cloud environment with Pulumi framework and Google Cloud Platform services. The sessions fits the best developers who want to learn and practice build and deploy techniques using Google Cloud Platform stack and Pulumi, Node.js, and NestJS.




Alex Korzhikov & Andrew Reddikh


repository-open-graph-template 1

Workshop Begins!



Who are we?

Alex Korzhikov

alex korzhikov photo

Software Engineer, Netherlands

My primary interest is self development and craftsmanship. I enjoy exploring technologies, coding open source and enterprise projects, teaching, speaking and writing about programming - JavaScript, Node.js, TypeScript, Go, Java, Docker, Kubernetes, JSON Schema, DevOps, Web Components, Algorithms 👋 ⚽️ 🧑‍💻 🎧

Andrew Reddikh

andrew reddikh photo

Software Engineer, United Kingdom

Passionate software engineer with expertise in software development, microservice architecture, and cloud infrastructure. On daily basis, I use Node.js, TypeScript, Golang, and DevOps best practices to build a better tech world by contributing to open source projects.

What are we going to do today?

⬆️ About

Which technologies are we going to use?

⬇️ Tags

Crypto 🦄 Currency Converter - Node.js

What we’re building

We’re building a currency converter, which can be used over gRPC calls.

currency convertor schema

Our intention is to send a request similar to convert 0.345 ETH to CAD and as a result we want to know the final amount in CAD and conversion rate. We also assume that, it could be more than one currency provider, e.g.

  1. Europe Central Bank rates
  2. Bank of England rates
  3. Crypto Rates

Here is how it works:


1. Checkout demo project

Let’s get started from cloning demo monorepo

git clone git@github.com:x-technology/micro-services-nestjs-infrastructure-pulumi-gcp.git

2. Install protoc

For efficient work with .proto format, and to be able to generate TypeScript-based representation of protocol buffers we need to install protoc library.

If you’re a MacOS user and have brew package manager, the following command is the easiest way for installation:

brew install protobuf
# Ensure it's installed and the compiler version at least 3+
protoc --version

For Linux users

Run the following commands:

curl -OL https://github.com/protocolbuffers/protobuf/releases/download/v3.14.0/$PROTOC_ZIP
sudo unzip -o $PROTOC_ZIP -d /usr/local bin/protoc
sudo unzip -o $PROTOC_ZIP -d /usr/local 'include/*'

Alternately, manually download and install protoc from here.

3. Prepare environment

Make sure we have Node.js v14+ installed. If not, nvm is a very good tool to install multiple node versions locally and easily switch between them.

Then we need to install dependencies within the monorepo.

npm install                 # turborepo and dependencies

Yay! 🎉 Now we’re ready to go with the project.

Monorepo structure

For better monorepo project management we used Turborepo

The project shapes into the following structure:

├── Dockerfile              # docker mono image
├── docker-compose.yaml     # docker-compose to start everything at once
├── docs                    # code examples
├── infrastructure          # pulumi and google cloud platform
├── node_modules            # common node modules
├── package.json            # project meta information & dependencies
├── packages                # microservices source code
└── turbo.json              # monorepo configuration

Let’s move on 🚚

Using Turborepo

Turborepo brings to the table generic configuration which can be easily applied across all/or filtered packages.

We use our common modules compiled to JavaScript, so before using it in services we need to build it first. The following commands executed build command against all common packages:

// ./package.json
"scripts": {
  "build": "turbo run build",
  "start": "turbo run start:prod"
// ./turbo.json
"pipeline": {
  "build": {
    "dependsOn": ["^build"],
    "outputs": ["dist/**"]
  "start:prod": {
    "cache": false
npm run build
npm run start
curl http://localhost:3001/currency-converter/convert

What is GRPC?

grpc logo aka pancakes

gRPC Remote Procedure Calls, of course! obvious reaction

gRPC is a modern, open source remote procedure call (RPC) framework that can run anywhere. It enables client and server applications to communicate transparently, and makes it easier to build connected systems

microservices graph


// http://protobuf-compiler.herokuapp.com/
syntax = "proto3";

package hello;

service HelloService {
  rpc JustHello (HelloRequest) returns (HelloResponse);

  rpc ServerStream(HelloRequest) returns (stream HelloResponse);

  rpc ClientStream(stream HelloRequest) returns (HelloResponse);

  rpc BothStreams(stream HelloRequest) returns (stream HelloResponse);

message HelloRequest {
  string greeting = 1;

message HelloResponse {
  string reply = 1;

client server communication

What are Protocol Buffers?

An efficient technology to serialize structured data

message Person {
  string name = 1;
  int32 id = 2;
  bool has_ponycopter = 3;

Does anyone know what numbers on the right side mean?

syntax = "proto3";

package hello;

service HelloService {
  rpc SayHello (HelloRequest) returns (HelloResponse);

message HelloRequest {
  string greeting = 1;

message HelloResponse {
  string reply = 1;

Demo - Hello Node.js GRPC

// package.json
"scripts": {
  "1. download prices": "node index.js",
  "2. generate protobuf runtime": "protoc --js_out=import_style=commonjs,binary:. prices.proto",
  "3. run protobuf transformation": "node index.js",
  "4. start grpc server": "node grpc-server.js",
  "5. start grpc client": "node grpc-client.js"

Demo - Run Microservices Locally

"workspaces": [
const all = require('@common/go-grpc')
const client = new all.ecbProvider.EcbProviderClient('', all.createInsecure());
const response = await client.GetRates(new all.currencyProvider.GetRatesRequest())
// inside converter container
const all = require('@common/go-grpc')
const client = new all.currencyConverter.CurrencyConverterClient('', all.createInsecure());
const response = await client.Convert(new all.currencyConverter.ConvertRequest({ sellAmount: 100, sellCurrency: 'USD', buyCurrency: 'GBP' }));
npm start

Nest Framework

Nest Icon

A progressive Node.js framework for building efficient, reliable and scalable server-side applications

// modules
import { Module } from '@nestjs/common';
import { CatsModule } from './cats/cats.module';

  imports: [CatsModule],
export class AppModule {}

// controllers
import { Controller, Get, Req } from '@nestjs/common';
import { Request } from 'express';

export class CatsController {
  findAll(@Req() request: Request): string {
    return 'This action returns all cats';

clean architecture

// @Body(), @Get(), @Post(), @Put(), @Delete(), @Patch(), @Options(), @Head(), @All()
findAll(@Req() request: Request): string {
  return 'This action returns all cats';

Demo - All Together

# speed run
npx @nestjs/cli new movies            # create new nestjs microservice
cd movies
npm run start

# or full log
rm -rf packages
rm -rf tmp
mv packages tmp
mv tmp packages
cd packages
mv packages monolith
npx @nestjs/cli new movies
cd movies
npm start

# root
cd ../../
git checkout demo-new-nestjs-module   # monorepo full example
npm install                           # run turborepo tasks
npm run build
npm start

# test
curl http://localhost:3000/
curl http://localhost:3001/

# see currency converter monolith project
curl http://localhost:3001/currency-converter/convert

Issues - Practical Exercises

1. Attach Swagger Plugin for each package and check content, is GRPC auto documented?

npm install --save @nestjs/swagger swagger-ui-express

2. Split packages/monolith into separate microservices.

Make sure turborepo configuration is updated, build and start all services

3. Fix type definitions inside controllers, why observable types are generated instead of Promises?

// packages/monolith/src/currency-converter/currency-converter.controller.ts
  async convert(): Promise<ConvertResponse> {
    // TODO fix type definitions
    const $rates = this.currencyConvertClient.Convert(ConvertRequest.fromJSON({
      sellAmount: 100,
      sellCurrency: 'USD',
      buyCurrency: 'GBP',

    const firstNumber = await firstValueFrom($rates as unknown as Observable<ConvertResponse>);
    const rates = firstNumber as unknown as ConvertResponse

    return rates

Infrastructure - Google Cloud Platform

Introduction to Google Cloud Platform

Google Cloud Platform is a enourmous cloud ecosystem that enables to organize, develop, publish applications worldwide

Docker - Build application as a docker image

Let’s have a quick look into Dockerfile and see how it works.

FROM node:16.18.0 as builder

RUN apt update && \
  apt install -y protobuf-compiler

# Create app directory
WORKDIR /usr/src/app

# A wildcard is used to ensure both package.json AND package-lock.json are copied
COPY package*.json turbo.json ./

COPY packages ./packages

# Install workspace dependencies
RUN npm install

# Creates a "dist" folder with the production build
RUN npm run build

# Start the server using the production build
CMD [ "node", "packages/dist/main.js" ]

GitHub Actions - Make builds automatically

Here is GitHub Actions file which defines building/publishing docker image on every push to repository in specific paths.

name: Docker Image Build

      - main
      - packages/**
      - Dockerfile
      - .dockerignore
    runs-on: ubuntu-latest
      - uses: actions/checkout@v2
      - name: Build the Docker image
          DOCKER_USERNAME: $\{\{ secrets.DOCKER_USERNAME \}\}
          DOCKER_PASSWORD: $\{\{ secrets.DOCKER_PASSWORD \}\}
        run: |
          IMAGE_TAGGED="xtechnology/$\{\{ github.event.repository.name \}\}:${GITHUB_SHA}"
          IMAGE_LATEST="xtechnology/\$\{\{ github.event.repository.name \}\}:latest"
          echo ${DOCKER_PASSWORD} | docker login -u ${DOCKER_USERNAME} --password-stdin
          docker build -t ${IMAGE_TAGGED} -t ${IMAGE_LATEST} .
          docker push ${IMAGE_TAGGED}
          docker push ${IMAGE_LATEST}
          docker logout



Pulumi - Developer-First Infrastructure as Code

Compared to Terraform?

Terraform is an open-source infrastructure as code software tool that provides a consistent CLI workflow to manage hundreds of cloud services.

Terraform resources

Pulumi and Google Cloud Platform (GCP) setup

Let’s get started from pulumi installation and initial infrastructure repo setup.

  1. Let’s first install gcloud with a command. Follow instructions here as installation depends on the machine and OS.

  2. Next, we need to get pulumi cli installed.
    brew install pulumi
  3. Now it’s time to create infrastructure project in infra folder with the following command.
    pulumi new typescript
  4. Let’s get authorized registered at GCP and Pulumi
    • https://console.cloud.google.com
    • https://app.pulumi.com
  5. Get authorized in cli commands:
    pulumi login
    gcloud login
    gcloud auth application-default login
  6. Get dependencies installed in infra folder
    npm install
  7. Configure access to GCP from pulumi
pulumi config set gcp:project <your-gcp-project-id> # e.g. xtechnology
pulumi config set gcp:region <your-region> # e.g us-west1
  1. Yay! Now we’re ready to start coding our infrastructure straight away.

Kubernetes Cluster

Let’s get started with Kubernetes Cluster in GCP, and for this purpose we’re going to use Pulumi to start.

We need to import a file, containing description of our cluster.

import { kubeconfig, cluster } from "./cluster";

export const clusterName = cluster.name;
export const config = kubeconfig;

Now, let’s try to do a simple command to build our infrastructure in the cloud:

pulumi up

Let’s check our cluster at the Google Cloud Platform. Great! It’s there, just in few lines of TypeScript code.

Install ingress into kubernetes cluster

What is Ingress? Ingress exposes HTTP and HTTPS routes from outside the cluster to services within the cluster. Traffic routing is controlled by rules defined on the Ingress resource.

Ingress chart

Let’s add it with the following code:

import { ingressServiceIP } from "./k8s/system";

export const externalIp = ingressServiceIP;

Right after applying this code, we can see externalIp it’s our public IP of the cluster. Now let’s attach DNS to this IP.


We’re going to use CloudFlare for DNS as it provides very rich api and also an extra features like Anti-Ddos and more.

import { mainRecord } from "./dns";

export const domain = mainRecord;

Once again pulumi up to see the changes applied.

Deploy NestJs Application

import { currencyConverter } from "./k8s/apps";

export const service = currencyConverter.urn;

Let’s find clusterName in pulumi output and use it to get access to cluster with kubectl command.

gcloud container clusters get-credentials apps-cluster-4eef4bd --region=europe-west2-b

Let’s create a proxy forwarding to our service inside the kubernetes cluster

kubectl port-forward -n apps-q0fg8ahd svc/currency-converter-grpc 8080:80

Now it’s the moment to call our currency-converter:

curl -X GET http://localhost:8080/currency-converter/convert

But, we have DNS, right. Let’s try the same call via domain name:

curl -X GET https://in1.app/currency-converter/convert

Great 🎉!

We’ve just created the full infrastructure and deployed our NestJs application into the Kubernetes cluster using Helm chart, pulumi, and Typescript.


Helm is a package manager for Kubernetes. Helm is the K8s equivalent of yum or apt. Helm deploys charts, which you can think of as a packaged application.

We store our helm charts inside the ./infra/charts folder.

By running a command we can create a new helm chart:

helm create grpc

Helm also provides an ability to easily template our package, so we can provide multiple values into the chart, when we deploy it.

Following command will show us rendered a yaml definition of the helm chart

helm template grpc


It’s time to have some practice and evolve our services even more!

Let’s grab a task based on the things you’d like to do 👇



Please share your feedback on our workshop. Thank you and have a great coding!

If you like the workshop, you can become our patron, yay! 🙏


microservices pulumi google cloud platform devops node.js nestjs javascript protobuf grpc typescript lerna npm yarn docker git architecture crypto currency