Using Mongoid with Ruby on Rails

Schema-less Design
An introduction to the Mongoid ODM

1. MongoDB
2. Mongoid ODM compares to ORM/SQL
3. Modelling Bitmaker Rainforest app without
ActiveRecord
4. Going further
5. Conclusion
Overview

Questions going in
• What is MongoDB?
• How does it compare from SQL-based databases?
• What does schema-less entail?
• How does ODM compare with ORM? Rainforest
tutorial
• Worth it?

Object Document Mapper (ODM)
Translate between objects in code and document
representation of data.
Object Relational Mapping translates between
objects in code the the relational representation of
the data.

MongoDB
• Open-source database
• NoSQL (cluster friendly, 21st-century web, non-
relational, schema-less)
• Data is stored in collections and documents
whereas in SQL data is stored in tables and rows
• JSON-like structure to documents

Differs from SQL
No JOINS -> run two or more queries
Embed and Referencing of document/data objects
Fast querying
PostgreSQL has strict schema / MongoDB has
implicit schema

Use cases
“Big Data” - large and unwieldy
90% of business data is unstructured
Increasing volume, variety and velocity
craigslist, Forbes, New York Times, Foursquare, etc.

The Setup
Install mongodb
https://docs.mongodb.org/manual/installation/
Run mongo server and mongo database in separate
terminal tabs

~>rails new rainforest —skip-active-record
The Setup
Removes ActiveRecord from the application
No schema.rb
No migrations folder
gem 'mongoid', '~> 5.1'
~>bundle install
~>rails generate mongoid:config
creates conﬁg/mongoid.yml

application.rb
require File.expand_path('../boot', __FILE__)
require "rails"
# Pick the frameworks you want:
require "active_model/railtie"
require "active_job/railtie"
# require "active_record/railtie"
require "action_controller/railtie"
require "action_mailer/railtie"
require "action_view/railtie"
require "sprockets/railtie"
require "rails/test_unit/railtie"
. . .

mongoid.yml
development:
# Configure available database clients. (required)
clients:
# Defines the default client. (required)
default:
# Defines the name of the default database that Mongoid can
connect to.
# (required).
database: rainforest_development
# Provides the hosts the default client can connect to. Must
be an array
# of host:port pairs. (required)
hosts:
- localhost:27017
options:
# Change the default write concern. (default = { w: 1 })
# write:
# w: 1

Rainforest App
class Product
include Mongoid::Document
field :name, type: String
field :description, type: String
field :price_in_cents, type: Integer
. . .
embeds_many :reviews
embeds_one :category
end
class Category
embedded_in :product
end

Product Controller
class ProductsController < ApplicationController
. . .
def product_params
params.require(:product).permit(:name, :description, :price_
in_cents, {category: [:category_id, :name]})
end
end

Rainforest App
class User
field :email, type: String
field :password, type: String
. . .
has_many :products
end
class Review
field :comment, type: String
field :user_id, type: BSON::ObjectId
. . .
embedded_in :product
end

Relations
Mongoid ODM associations similar to ActiveRecord ORM
Association is not just for Relational Mappers
Embeds One, Embeds Many, Has One, Has Many, Has And
Belongs To Many, counter_cache (cache the counts of an
associated model)
No has_many :through

BSON Objects
BSON: supports embedding of objects within arrays
Allow for extensions not part of JSON spec
Minimum overhead
Traverse quickly (fast to encode and decode)

{
"_id" : ObjectId("5728b57b370d93124e000000"),
"name" : "Day planner",
"description" : "Schedule events",
"price_in_cents" : 1299,
"user_id" : ObjectId("5727b3b0370d93009d000000"),
"category" : {
"_id" : ObjectId("5728b57b370d93124e000001"),
"name" : “Office material"
},
"reviews" : [
{
"_id" : ObjectId("5728b5a3370d93124e000002"),
"comment" : "A great product",
"user_id" : ObjectId("5727b3b0370d93009d000000")
}

[1] pry(main)> Product.first
#<Product:0x007fc1290cd108> {
:_id => BSON::ObjectId('56b54bc3370d93d2d5411757'),
:category => {
"_id" => BSON::ObjectId('570c78f1370d936468000000'),
"name" => "Electronics"
},
:description => "Listen to music without all the hassle of a cord.",
:name => "Wireless Headphones",
:price_in_cents => 60,
:reviews => [
[0] {
"_id" => BSON::ObjectId('56b54e59370d93d3b4fe9848'),
"comment" => "Test comment”,
"user_id" => BSON::ObjectId('56eb9d92370d937950d21463
},
[1] {
"_id" => BSON::ObjectId('56b54e5f370d93d3b4fe9849'),
"comment" => "Checking commentrn”,
},
[2] {
"_id" => BSON::ObjectId('56b54e6d370d93d3b4fe984a'),
"comment" => "Just bought these headphones.”,
},
[0] {
"_id" => BSON::ObjectId('5706dc66370d934ccd000000'),
"comment" => "Checking embedded review",
"user_id" => BSON::ObjectId('56eb9d92370d937950d21463')
}
]
}

Rails Console
[3] pry(main)> products = Product.all
#<Mongoid::Criteria
selector: {}
options: {}
class: Product
embedded: false>
The query doesn’t run until the data is requested

[6] pry(main)> products.each {|product|puts product.name }
Wireless Headphones
Chromecast for TV
Ferrari
ATM
Computer
Ferrari2
Game Boy
Sony Headphones
Speakers
Car
Dell Computer
Samsung Television
Audi Convertible
Snowmobile
#<Mongoid::Contextual::Mongo:0x007fc129167550 @cache=nil, @klass=Product,
@criteria=#<Mongoid::Criteria
selector: {}
options: {}
class: Product
embedded: false>
, @collection=#<Mongo::Collection:0x70233797373000
namespace=rainforest_development.products>, @view=#<Mongo::Collection::View:0x70233797372780
namespace='rainforest_development.products @selector={} @options={}>, @cache_loaded=true>

Implicit schema
[9] pry(main)> product[:model] = "TS2-3Q78"
[8] pry(main)> product = Product.first

[1] pry(main)> Product.first
#<Product:0x007fc1290cd108> {
:_id => BSON::ObjectId('56b54bc3370d93d2d5411757'),
:category => {
"_id" => BSON::ObjectId('570c78f1370d936468000000'),
"name" => "Electronics"
},
:description => "Listen to music without all the hassle of a cord.",
:model => "TS2-3Q78",
:name => "Wireless Headphones",
:price_in_cents => 60,
:reviews => [
[0] {
"_id" => BSON::ObjectId('56b54e59370d93d3b4fe9848'),
"comment" => "Test comment"
},
[1] {
"_id" => BSON::ObjectId('56b54e5f370d93d3b4fe9849'),
"comment" => "Checking commentrn"
},
[2] {
"_id" => BSON::ObjectId('56b54e6d370d93d3b4fe984a'),
"comment" => "Just bought these headphones."
},
[3] {
"_id" => BSON::ObjectId('5706dc66370d934ccd000000'),
"comment" => "Checking embedded review",
"user_id" => BSON::ObjectId('56eb9d92370d937950d21463')
}
]
}

Mongoid
Similar syntax to ActiveRecord - associations, datatypes, etc.
No need for migrations
Great documentation; active community
Popular gems have mongoid counterpart (Devise,
carrierwave-mongoid, mongoid-paperclip, geocoder,
mongoid-rspec)
Setup is fast and mistakes quickly fixed (no rollbacks)

class Product
include Mongoid::Paperclip
. . .
has_mongoid_attached_file :avatar, styles: { medium: "300x300>",
thumb: "100x100>" }, default_url: lambda { |image|
ActionController::Base.helpers.asset_path('missing.png') }
validates_attachment_content_type :avatar,:content_type =>
["image/jpg",
"image/jpeg",
"image/png",
"image/gif"]
end
class ProductsController < ApplicationController
. . .
def product_params
params.require(:product).permit(:name, :description, :price_in_cents,
:avatar, {category: [:category_id, :name]})
end

#<Product:0x007fc12b0db148> {
:_id => BSON::ObjectId('570c32f9370d935b78000000'),
:avatar_content_type => "image/jpeg",
:avatar_file_name => "snowmobile.jpeg",
:avatar_file_size => 7081,
:avatar_fingerprint => "33234414813b3e3bf8a26281a2d7316f",
:avatar_updated_at => 2016-04-12 00:56:31 UTC,
:category => {
"_id" => BSON::ObjectId('570c32f9370d935b78000001'),
"name" => "Automobile"
},
:description => "For northern living",
:name => "Snowmobile",
:price_in_cents => 345436
}

Going further
• Indexing
• Map/Reduce to condense large volumes of data; work
around for JOINS - rearrange data into aggregate
forms
• Sharding
• Track data versioning
• GeoSpatial indexing

Worth it?
• It depends what you want
• Have MongoDB serve up user info or cache preferences, integrate with social
networks; SQL to handle payments/ordering
• Great for searching large amounts of data.
• PostgreSQL JSON type - support for arbitrary attributes (WARNING: anti-pattern)
• For this application, mongoid can work and clarified features for me, but SQL
and Schema-based design seem optimized for this case
• Different ways to organize data and design models
• Learned some new things about ActiveRecord and PostgreSQL

Sources
https://www.mongodb.com/nosql-explained
Rege, Gautam. Ruby and MongoDB Web Development Beginner’s Guide, Packt
Publishing, 2012
http://www.sarahmei.com/blog/2013/11/11/why-you-should-never-use-mongodb/
comment-page-1/
https://www.mongodb.com/compare/mongodb-mysql
https://github.com/meskyanichi/mongoid-paperclip
https://gorails.com/guides/setting-up-rails-4-with-mongodb-and-mongoid
https://gorails.com/blog/rails-4-0-with-mongodb-and-mongoid
http://stackoverflow.com/questions/26182890/perform-atomic-block-transactions-in-
rails-with-mongoid
http://blog.2ndquadrant.com/postgresql-anti-patterns-unnecessary-jsonhstore-
dynamic-columns/

@MrNickAltobelli
nmhalt@gmail.com
github: DataNick
www.nickaltobelli.com
Thank You!

Using Mongoid with Ruby on Rails

Recommended

More Related Content

What's hot (7)

Similar to Using Mongoid with Ruby on Rails (20)

Recently uploaded (20)

Using Mongoid with Ruby on Rails