# Installation

In this chapter, we will cover :

# 0. Introduction

So, you are interested in Storefront API? Well, that's why you are here. You've made a choice. Your decision will most certainly pay off, which is great. Be it developers, entrepreneurs, or even marketing managers who may want to try something new with the hope of enhancing their clients' or customers' experience, you chose the right path. We will explore everything you need to get you started with Storefront API infrastructure.

# 1. Installation with Docker

Docker has been arguably the most sought-after development tool ever brought to market, and has taken took the community by storm. Although it's still controversial whether it's the best choice among its peers, I have never seen such unanimous enthusiasm over one tech product throughout the whole developer community.

Why is this so? In modern computer engineering, products are so complex, with an endless list of dependencies intertwined with each other. Building such dependencies in place, for every situation where they are required, is one hell of a job, to say nothing of the glitches from all the version variations. That's where Docker steps in to help you achieve infrastructure automation. This concept was conceived to help you focus on your business logic rather than having you stuck with the hassles of lower-level tinkering.

Luckily, we have already been through all this for you, getting our hands dirty and doing a lot of the legwork for you. All you need to do is run a set of docker commands to get you up and running from scratch. Without further ado, let's get started!

# 1. Preparation

• You need docker and docker-compose installed.

• You need git installed.

# 2. Recipe

1. First, download Storefront API from github.

git clone https://github.com/DivanteLtd/storefront-api.git storefront-api
cd storefront-api

2. Copy ./config/default.json to ./config/local.json

cp config/default.json config/local.json

Then edit local.json to suit your needs.

3. Run the following Docker command :

To use Storefront API with embedded Elastic 7:

docker-compose -f docker-compose.yml up -d

Then, to restore the demo data set, please run: docker exec -it sfa_app_1 yarn restore7

The result would look something like this :

Building app
Step 1/8 : FROM node:10-alpine
 ---> 9dfa73010b19
Step 2/8 : ENV VS_ENV prod
 ---> Using cache
 ---> 4d0a83421665
Step 3/8 : WORKDIR /var/www
 ---> Using cache
 ---> e3871c8db7f3
Step 4/8 : RUN apk add --no-cache curl git
 ---> Using cache
 ---> 49e996f0f6cb
Step 5/8 : COPY package.json ./
 ---> 14ed18d76efc
Step 6/8 : RUN apk add --no-cache --virtual .build-deps ca-certificates wget &&     yarn install --no-cache &&     apk del .build-deps
 ---> Running in 3d6f91acc2fe
fetch http://dl-cdn.alpinelinux.org/alpine/v3.9/main/x86_64/APKINDEX.tar.gz
fetch http://dl-cdn.alpinelinux.org/alpine/v3.9/community/x86_64/APKINDEX.tar.gz
(1/2) Installing wget (1.20.3-r0)
(2/2) Installing .build-deps (0)
Executing busybox-1.29.3-r10.trigger
OK: 22 MiB in 26 packages
yarn install v1.16.0
info No lockfile found.
[1/4] Resolving packages...
warning @babel/node > @babel/polyfill@7.4.4: 🚨 As of Babel 7.4.0, this
package has been deprecated in favor of directly
including core-js/stable (to polyfill ECMAScript
features) and regenerator-runtime/runtime
(needed to use transpiled generator functions):

  > import "core-js/stable";
  > import "regenerator-runtime/runtime";
warning eslint > file-entry-cache > flat-cache > circular-json@0.3.3: CircularJSON is in maintenance only, flatted is its successor.
[2/4] Fetching packages...

# ... abridged

3. In order to verify, run docker ps to show which containers are up:

docker ps 

Then,

CONTAINER ID        IMAGE                     COMMAND                  CREATED             STATUS              PORTS                                            NAMES
53a47d5a6440        sfa_kibana   "/bin/bash /usr/loca…"   31 seconds ago      Up 29 seconds       0.0.0.0:5601->5601/tcp                           sfa_kibana_1
7d8f6328601b        sfa_app      "docker-entrypoint.s…"   31 seconds ago      Up 27 seconds       0.0.0.0:8080->8080/tcp                           safa_app_1
165ae945dbe5        sfa_es1      "/bin/bash bin/es-do…"   8 days ago          Up 30 seconds       0.0.0.0:9200->9200/tcp, 0.0.0.0:9300->9300/tcp   elasticsearch
8dd144746cef        redis:4-alpine            "docker-entrypoint.s…"   11 days ago         Up 31 seconds       0.0.0.0:6379->6379/tcp                           sfa_redis_1

The ports number will be used later in the frontend configuration. In fact, they are already set as default values.

You will see four containers are running, which are:

# 3. Peep into the kitchen (what happens internally)

We used docker-compose to set up the entire environment of Storefront API. It was more than enough to launch the machines needed behind the scenes to run the shop.

It was possible because docker encapsulated the whole infrastructure into a linear set of declarative descriptions for the desired state.

The docker-compose command took a yml file for input. This file describes its base requirements, but also Storefront API itself; that is, Elasticsearch as a data store, Redis for caching and Kibana for helping you grab your data visually (a partner to Elasticsearch).

version: '3.0'
services:
  es1:
    container_name: elasticsearch
    build: docker/elasticsearch/
    ulimits:
      memlock:
        soft: -1
        hard: -1    
    volumes:
      - ./docker/elasticsearch/config/elasticsearch.yml:/usr/share/elasticsearch/config/elasticsearch.yml:ro
    ports:
      - '9200:9200'
      - '9300:9300'
    environment:
      - discovery.type=single-node
      - cluster.name=docker-cluster
      - bootstrap.memory_lock=true
      - "ES_JAVA_OPTS=-Xmx512m -Xms512m"

  kibana:
    build: docker/kibana/
    volumes:
      - ./docker/kibana/config/:/usr/share/kibana/config:ro
    ports:
      - '5601:5601'
    depends_on:
      - es1

  redis:
    image: 'redis:4-alpine'
    ports:
      - '6379:6379'
  app:
    # image: divante/storefront-api:latest
    build:
      context: .
      dockerfile: docker/storefront-api/Dockerfile
    depends_on:
      - es1
      - redis
    env_file: docker/storefront-api/default.env
    environment:
      VS_ENV: dev
    volumes:
      - './config:/var/www/config'
      - './ecosystem.json:/var/www/ecosystem.json'
      - './migrations:/var/www/migrations'
      - './package.json:/var/www/package.json'
      - './babel.config.js:/var/www/babel.config.js'
      - './tsconfig.json:/var/www/tsconfig.json'
      - './nodemon.json:/var/www/nodemon.json'
      - './graphql-schema-linter.config.js:/var/www/graphql-schema-linter.config.js'
      - './scripts:/var/www/scripts'
      - './src:/var/www/src'
      - './var:/var/www/var'
    tmpfs:
      - /var/www/dist
    ports:
      - '8080:8080'

      

volumes:
  esdat1:

version denotes which version of docker-compose this file uses.

services describe containers and codifies how they should run. In other words, it codifies option flags used with docker run ...

es1 contains information about the data store Elasticsearch container.

• build denotes the build path of the container.
• volumes contains the mount paths of volumes shared between host and container, defined as host:container
• ports connects ports between the host and container, defines as host:container
• environment allows you to add environment variables. Xmx512m means JVM will take up to a maximum of 512MB memory. Xms512m means minimum memory. Combining them, there will be no memory resize, it will just stick to 512MB from start to finsih throughout its life cycle.

kibana contains information about the Kibana application container.

• depends_on defines dependencies of a container on other containers. So, this container is dependent on es1 which was described above.
• volumes means volumes shared, :ro creates the volume in read-only mode for the container.

redis contains information about the Redis cache application container.

• The image node contains the name of the image this container is based on.

volumes can be defined at the top level to as a reference to be used across multiple services (containers).

app contains information about the Storefront API application.

• build is the path for build information. If the value is string, it's a plain path. When it's an object, you may have a few options to add. context is a relative path or git repo url where the Dockerfile is located. The dockerfile node may change the path/name of Dockerfile. more info
• depends_on tells us this container is based on the es1 and redis containers we created above.
• env_file helps you add environment values from files. It's a relative path from the docker-compose file that is in the process. In this case, it's docker-compose.nodejs.yml
• environment is used to set VS_ENV as dev so that the environment will be set up for developer mode.
• tmpfs denotes temporary volumes that are only available to host memory. Unlike volumes, this tmpfs will be gone once the container stops. This option is only available on Linux.

# 2. Install with npm packages

1. Install all needed packages @storefront-api/core, @storefront-api/default-vsf, @storefront-api/default-catalog, @storefront-api/default-img, @storefront-api/platform-magento2.

yarn add @storefront-api/core@1.0.0-rc.2 @storefront-api/default-vsf@1.0.0-rc.2 @storefront-api/default-catalog@1.0.0-rc.2 @storefront-api/default-img@1.0.0-rc.2 @storefront-api/platform-magento2@1.0.0-rc.2

2. After that is done you only need to create an config folder and put all the contend of storefront-api/tree/master/config in there. When that is done you just need to copy the contend of code example into your index.js.

// index.js
const { Server } = require('@storefront-api/core')
const { DefaultVuestorefrontApiModule } = require('@storefront-api/default-vsf')
const { DefaultCatalogModule } = require('@storefront-api/default-catalog')
const { DefaultImgModule } = require('@storefront-api/default-img')
const magento2 = require('@storefront-api/platform-magento2')

 let modules = [
    DefaultVuestorefrontApiModule({
        platform: {
            name: 'magento2',
            platformImplementation: magento2
        }
    }),
    DefaultCatalogModule(),
    DefaultImgModule(),
]

const server = new Server({
    modules
})

server.start();

3. Now you only need to run node index.js