pokemon are cool

Mongo as a Docker container

run in docker container and expose mongo port 27017 on localhost docker run -d --rm --hostname mongo --name mongo -p 27017:27017 mongo

ways to connect:

• GUI
• programmatic via a Driver (e.g. Nodejs driver https://mongodb.github.io/node-mongodb-native/)
• cli access (MongoDb shell)

• docker exec -it <CONTAINER_ID> /bin/bash then mongo
• or directly docker exec -it <CONTAINER_ID> mongo

• physical database, i.e. the mongodb server
• logical databases (db) - e.g. for each application/service a different database
• collections - e.g all users
• namespace = a logical db + a collection
• documents - e.g. a single user

Cheatsheet

show dbs
use pokemonDB
db.createCollection('pokemon')
db.createCollection('trainers')
show collections
db.trainers.find({name:"Ashe"})
db.trainers.findOne({name:"Ashe"})
db.trainers.update({name:"Ashe"},{$set:{level:7}})
db.trainers.update({name:"Ashe"},{name:"Ash Clown"})
const t = db.trainers
t.update({name:"Misty"},{},{upsert:true})
t.deleteMany({name:{$exists:false}})
t.deleteOne({_id:ObjectId("6073170842691fc14f93caa4")})

• CRUD
• no-sql vs sql based Dbs
• https://en.wikipedia.org/wiki/Database_normalization
• schema for mongo -> mongoose
• Mongo Queries
• MongoDb NodeJS driver
• MongoDb NodeJS API docs
• TypeScript Starter Template

• Indices - make things fast
• Aggregation Pipelines - array.reduce() for lists of documents - can also be done completely on application level (TS/JS), but DB is probably faster