Exploring GraphQL Systems Lifecycle: Server, Development, Monitoring, and Maintenance

GraphQL Systems Lifecycle: Server, Development, Monitoring, and Maintenance
- TODO Deck, Design, Colors, and Images
About Town (Core Concepts)
The Trail (REST, aysnc, docs)
The Ascent (Development Practices)
Going to the People (API Contract, Monitoring, CI/CD)
Produced for Inspection (Performance and Monitoring)
The First Day of Creation (Cloud and Serverless)
- GAE
- Heroku
- Serverless
- Cloud
- AWS AppSync
- Google Run
- Azure Functions
- Firebase

GraphQL Systems Lifecycle: Server, Development, Monitoring, and Maintenance

https://www.meetup.com/bostonpython/events/262475756/

Now, what do you do with it? Jason Walsh (Senior Engineer, Chewy) continues our fable with the ongoing cycle of software from the server's perspective. State, async, fetching, oh my! As well as some basic implementations, he'll also cover things like tracing, logging, validation and the ongoing maintenance cycle of api.

Provide background on historical means of separating concerns
Contrast GraphQL with REST
Focus on hosted Python
Indicate design considerations
Managing change and code ownership
Focus on exploration
Look at extending core cases
What how do you detect issues with resolvers
Performance considerations over the wire

TODO Deck, Design, Colors, and Images

GraphViz
Core presentation equipement
Code repository for examples

About Town (Core Concepts)

Evaluate and Assess GraphQL for your Organization

Example: Existing REST microservices and multiple clients

Take a simple example and walk through some of the considerations associated with design, development, deployment, change management,

Show a high level for an example running locally.

Getting Started and Terminology

Python and GraphQL

https://graphene-python.org/

import graphene

class Query(graphene.ObjectType):
    hello = graphene.String()

    def resolve_hello(self, info):
        return 'World'


schema = graphene.Schema(query=Query)

schema.execute('''
  query {
    hello
  }
''')

GraphQL queries, mutations, and subscriptions

Type Definition

Resolvers

Faker
Random
Neo4J
DB

Schema

type Query {
    get(id: ID, meta: String): Thing
}

type Thing {
    id: ID!
    title: String!
    meta: String
}

schema {
    query: Query
}

Executable Schema

Model

https://medium.com/google-cloud/secure-graphql-apis-in-minutes-with-google-cloud-run-and-grand-stack-97d050dbc744

type Person {
   name: String!
   knows: [Person] @relation(name: "KNOWS", direction: "OUT")
   likes: [Hobby] @relation(name: "LIKES", direction: "OUT")
}
type Hobby {
   name: String!
   liked_by: [Person] @relation(name: "LIKES", direction: "IN")
}

The Trail (REST, aysnc, docs)

What does GraphQL get you that REST doesn't and how we're transitioning from server-side template driven develoment to

Assuming we have the following industry adoption timeline.

Workflow with REST

As an alternative to server-side rendering and

Workflow with Microservices

Workflow with API Gateways

Weak Typing

Async Stitching

REST Primitives

Documentation and OpenAPI

Subscriptions

Round Trip Times

Getting Data

At the most simple the option here is to ensure that a client can get data for the request. Since we're pushing

State

There are three core components when making requests to the GraphQL end-point: the core mutation or query, parameters passed into that document, and the headers associated with the request.

Async

API Gateways

The Ascent (Development Practices)

https://graphql.org/learn/best-practices/

Design/Schema First vs. Code First

Ownership Consolidation

Designing API

https://docs.aws.amazon.com/appsync/latest/devguide/designing-a-graphql-api.html

build a schema from scratch, provision resources automatically, manually define a data source, and connect to it with a GraphQL resolver

Create Schema

IBM OpenAPI-to-GraphQL

https://www.ibm.com/blogs/research/2019/07/openapi-graphql/

Security

Authentication

cookies, JSON web tokens, or even HTTP Basic auth.

Authorization

From REST

https://blog.logrocket.com/how-to-secure-a-rest-api-using-jwt-7efd83e71432/

From Session Cookies

https://stackoverflow.com/questions/51977753/apollo-2-0-0-graphql-cookie-session

Basic Auth

String usernameColonPassword = "user:passwd";
String basicAuthPayload = "Basic " + Base64.getEncoder().encodeToString(usernameColonPassword.getBytes());

Going to the People (API Contract, Monitoring, CI/CD)

Exploration

Data

Types

Specifications

Errors

Monitoring Usage

Breaking Changes

Deprecation Strategies

Mocking and Testing

There are a number of excellent resources

https://github.com/graphql/swapi-graphql

Expanding the Team

Managing Teams and Federation

Federated graph

GrAMPS

https://gramps-graphql.github.io/gramps-express/

https://about.sourcegraph.com/graphql/graphql-at-massive-scale-graphql-as-the-glue-in-a-microservice-architecture

CI/CD and Tooling

Produced for Inspection (Performance and Monitoring)

Performance Response Payload

Tracing

https://github.com/graphql-python/graphene-tornado/pull/34

This would cover some of the core operational.

Graphene Core

Installation

Tracing Middleware

https://docs.graphene-python.org/en/latest/execution/

Performance Characteristics

https://github.com/apollographql/apollo-tracing

The First Day of Creation (Cloud and Serverless)

https://www.apollographql.com/platform/

Cloud

AWS AppSync

AWS AppSync is an enterprise level, fully managed GraphQL service with real-time data synchronization and offline programming features.

Google Run