How to deploy Super Graph

Since you're reading this you're probably considering deploying Super Graph. You're in luck it's really easy and there are several ways to choose from. Keep in mind Super Graph can be used as a pre-built docker image or you can easily customize it and build your own docker image.

JWT tokens (Auth0, etc)

When deploying on a subdomain and configure this service to use JWT authentication. You will need the public key file or secret key. Ensure your web app passes the JWT token with every GraphQL request (Cookie recommended). You have to add the web domain to the cors_allowed_origins config option so CORS can allow the browser to do cross-domain ajax requests.

Google Cloud Run (Fully Managed)

Cloud Run is a fully managed compute platform for deploying and scaling containerized applications quickly and securely. Your Super Graph app comes with a cloudbuild.yaml file so it's really easy to use Google Cloud Build to build and deploy your Super Graph app to Google Cloud Run.

note

Remember to give Cloud Build permission to deploy to Cloud Run first this can be done in the Cloud Build settings screen. Also the service account you use with Cloud Run must have the IAM permissions to connect to CloudSQL. https://cloud.google.com/sql/docs/postgres/connect-run

Use the command below to tell Cloud Build to build and deploy your app.

gcloud build submit --substitutions=SERVICE_ACCOUNT=admin@my-project.iam.gserviceaccount.com,REGION=us-central1 .
Secrets Management

Your secrets like the database password should be managed by the Mozilla SOPS app. This is a secrets management app that encrypts all your secrets and stores them in a file to be decrypted in production using the Cloud KMS (Google Cloud KMS Or Amazon KMS). Our cloud build file above expects the secrets file to be config/prod.secrets.yml. You can find more information on Mozilla SOPS on their site. https://github.com/mozilla/sops

Build Docker Image Locally

If for whatever reason you decide to build your own Docker images then just use the command below.

docker build -t your-api-service-name .

With a Rails app

Super Graph can read Rails session cookies, like those created by authentication gems (Devise or Warden). Based on how you've configured your Rails app the cookie can be signed, encrypted, both, include the user ID or just have the ID of the session. If you have choosen to use Redis or Memcache as your session store then Super Graph can read the session cookie and then lookup the user in the session store. In short it works really well with almost all Rails apps.

For any of this to work Super Graph must be deployed in a way that make the browser send the apps cookie to it along with the GraphQL query. That means Super Graph needs to be either on the same domain as your app or on a subdomain.

I need an example

Say your Rails app runs on myrailsapp.com then Super Graph should be on the same domain or on a subdomain like graphql.myrailsapp.com. If you choose subdomain then remeber read the Deploy under a subdomain section.

Deploy under a subdomain

For this to work you have to ensure that the option :domain => :all is added to your Rails app config Application.config.session_store this will cause your rails app to create session cookies that can be shared with sub-domains. More info here /sharing-a-devise-user-session-across-subdomains-with-rails

With NGINX

If your infrastructure is fronted by NGINX then it should be configured so that all requests to your GraphQL API path are proxyed to Super Graph. In the example NGINX config below all requests to the path /api/v1/graphql are routed to wherever you have Super Graph installed within your architecture. This example is derived from the config file example at /microservices-nginx-gateway/nginx.conf

NGINX with sub-domain

Yes, NGINX is very flexible and you can configure it to keep Super Graph a subdomain instead of on the same top level domain. I'm sure a little Googleing will get you some great example configs for that.

# Configuration for the server
server {
# Running port
listen 80;
# Proxy the graphql api path to Super Graph
location /api/v1/graphql {
proxy_pass http://super-graph-service:8080;
proxy_redirect off;
proxy_set_header Host $host;
proxy_set_header X-Real-IP $remote_addr;
proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
proxy_set_header X-Forwarded-Host $server_name;
}
# Proxying all other paths to your Rails app
location / {
proxy_pass http://your-rails-app:3000;
proxy_redirect off;
proxy_set_header Host $host;
proxy_set_header X-Real-IP $remote_addr;
proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
proxy_set_header X-Forwarded-Host $server_name;
}
}

On Kubernetes

If your Rails app runs on Kubernetes then ensure you have an ingress config deployed that points the path to the service that you have deployed Super Graph under.

Ingress config

apiVersion: extensions/v1beta1
kind: Ingress
metadata:
name: simple-rails-app
annotations:
nginx.ingress.kubernetes.io/rewrite-target: /
spec:
rules:
- host: myrailsapp.com
http:
paths:
- path: /api/v1/graphql
backend:
serviceName: graphql-service
servicePort: 8080
- path: /
backend:
serviceName: rails-app
servicePort: 3000

Service and deployment config

apiVersion: v1
kind: Service
metadata:
name: graphql-service
labels:
run: super-graph
spec:
ports:
- port: 8080
protocol: TCP
selector:
run: super-graph
---
apiVersion: apps/v1
kind: Deployment
metadata:
name: super-graph
spec:
selector:
matchLabels:
run: super-graph
replicas: 2
template:
metadata:
labels:
run: super-graph
spec:
containers:
- name: super-graph
image: docker.io/dosco/super-graph:latest
ports:
- containerPort: 8080