graphql-subscriptions VS serverless-graphql

graphql-yoga is built on other packages that provide functionality required for building a GraphQL server such as web server frameworks like express and apollo-server, GraphQL subscriptions with graphql-subscriptions and subscriptions-transport-ws, GraphQL engine & schema helpers including graphql.js and graphql-tools, and an interactive GraphQL IDE with graphql-playground.

The most common used (but not best maintained) library for such a PubSub engine in the GraphQL context is graphql-subscriptions. There are also adapters available for more distributed systems (where all GraphQL API replicas must be notified about the event) e.g. over Redis.

During that work, we created and merged the reference implementation into graphql-js and created two supporting libraries: graphql-subscriptions and subscriptions-transport-ws. Here is a talk with deep dive into all the details.

The Prisma / GraphQL-Yoga comes with Subscriptions, and this was the first thing I came across when my client asked for realtime updates. The implementation was quite difficult - took me a long time. Eventually, it was working locally, and in staging. However, when it came to the production environment, for some reason it just didn't work! This unfortunately ended up with a user losing an auction, which ultimately went to court etc.

There are also other solutions available: https://github.com/apollographql/graphql-subscriptions#pubsub-implementations

First, let's try to understand what PubSub is exactly. Apollo Server uses a publish-subscribe (pub/sub) model to track events that update subscriptions. The graphql-subscriptions library included in all apollo-server packages (including middleware integrations) provides a PubSub class as a basic in-memory event bus.

To add this, I can just expand the basic setup above. To do so, I do actually use a couple of small Apollo modules! Most can be picked and configured independently. For this case, graphql-subscriptions provides a little bit of pubsub logic that works within resolvers, and subscriptions-transport-ws integrates that into Express to handle the websockets themselves. Super helpful

Since then, the ecosystem has changed. Using the Serverless Framework makes deployment simpler. We released the managed Tracetest App making any serverless-based systems simpler to instrument and test. You can now test public-facing apps with no infra overhead!

We see some great results from using these in conjunction with frameworks such as SST or Serverless, and also some real spaghetti from people who organically proliferate 100’s of functions over time and lose track of how they relate to each other or how to update them safely across time and service. Buyer beware!

When we think about AWS serverless service, the first thing that comes to our mind is Lambda function. Yes, the quickest way to deploy this backend Express JS app to AWS is to deploy it as a Lambda function. The easiest way is using Serverless Framework.

Serverless Framework: https://www.serverless.com/

Github | Website

In this example, we're using the serverless framework to quickly set up the Lambda function along with an API gateway for the entry point. The lambda function is a simple Koa REST API with a few functional endpoints.

Naturally, there are several options available to declare your cloud resources. The options with the most popularity are the CDK, AWS CloudFormation, SST, Serverless framework, Terraform, and AWS SAM. There are others, but when talking about Infrastructure as Code (IaC), these are the ones you hear about most often.

1 - Serverless + AWS CDK + Lift: An integration that amps up the traditional Serverless Framework with Lift's static frontend construct and CDK's robust infra definition.

It's a bit easier in Python if you use tools like https://www.serverless.com/. I'm not sure if Rust has something similar yet.

Serverless