///

Single API using Koa, apollo-server-koa and sequelize with postgres dialect

Before starting

Create database songkick
create a role
create the (tune) tables + partitions

By using the sql script in the root of project.

For ease, a .devcontainer directory for those who use vscode IDE is also created

Why Postgresql:

Partition data
Monitoring and tune top queries (pg_activity)
using copy (very fast) for insertion and queries
Postgis extension for spatial data

Requirements

Node.js
Docker
Docker-compose
Vscode (remote container extension)

Usage

 In Vscode: Remote-Containers: Rebuild Container

This create needed container

container App: all dev requirements: Node;js Postgres-client(psql)
container db: Postgres server 13

Execute sql script In the App container

psql -h db -U postgres -f sql/init.sql

load inital data

yarn && yarn loadData

Run the server

yarn start

Server will be running at http://localhost:3000/graphql

EndPoint

With Postman or curl

curl -XPOST http://localhost:3000/graphql \                 
> -H "Content-Type: application/json" \
> -d '{"query": "{concertsByBandAndLocationWithInRadius(bandIds: [129, 95, 123, 135] latitude: 48.8598659 longitude: 2.347761 radius: 50) { venue {id name latitude longitude }}}"}'

Description

For data queries, graphql is used to fetch data that we need only

Data are stored in postgresql database, for performance purpose partition are created by range of data for concerts table.

Postgis extension is installed, for effeciecy, indexing geolocation data, location field with a type geography is added to the table venue. we combine the longitude and latitude to create Point for each venue. A lot of functions to handle spatial data are provided by the extension ex: ST_DWithin ST_Distance.

nadhub / backend

///