![MONGOID
User.any_of(
{ :name => ‘John Doe’ },
{ :money.gt 10 }
)
User.any_in(:name => [‘John’, ‘Foo’, ‘Bar’])
User.where(:name => ‘John Doe’).and(:money.gt 10)
Post.all_in(:tags => [‘ruby’, ‘code’, ‘mongo’])](https://image.slidesharecdn.com/nosql-redis-mongo-110218214035-phpapp02/85/Nosql-redis-mongo-24-320.jpg)
Nosql redis-mongo
- 2. SQL: THE BEGINNING • Data structured on tables using columns • Data on one column is the key for making relationships with other tables • Normalization • RDBMS
- 3. SQL EXAMPLES
- 4. RDBMS BENEFITS • Normalization (a common standard format for accessing data) • Data consistency • Persistence • Consistency • Replication
- 6. RDBMS PROBLEMS • Highcost on performing complex operations over a big database • High system load: scalability needs • Difficult failover recovery • Difficult denormalization
- 7. NOSQL TO THE RESCUE!
- 8. NOSQL IS... •A storage system for format free data • Web scale • Software darwinism • Different types of solutions depending on the needs: Key-value, Documents, Graphs, Objects...
- 9. REDIS • http://redis.io Redis is SEXY • Key-value store • In memory • Persistent (periodical asynchronous writes to disk) • All operations are atomic
- 10. REDIS Types of data Cool operations • Hashes • RPOPLPUSH • Sets (arrays) • Pub/Sub • Lists
- 11. REDIS HSET myuser “email” “ivan@ivanhq.net” HSET myuser “web” “http://ibelmonte.com” HSET myuser “blog” “http://ivanh.net” HGET myuser “email” - ivan@ivanhq.net
- 12. REDIS sadd myuser:books “jackass” sadd myuser:books “jackass2” sadd myuser:books “Dude, where’s my car?” sadd myuser:books “The hangover” smembers myuser:books 1. jackass 2. jackass2 3. Dude, where’s my car? 4. The hangover
- 13. REDIS rpush friends “John Doe” rpush friends “Foo” rpush friends “Bar” lrange friends 0 -1 1. John Doe 2. Foo 3. Bar
- 14. REDIS rpush nonfriends “Peter Pan” rpush nonfriends “Batman” rpush nonfriends “Betty Boop” lrange nonfriends 0 -1 1. Peter Pan 2. Batman 3. Betty Boop
- 15. REDIS rpoplpush nonfriends friends lrange friends 0 -1 1. John Doe 2. Foo 3. Bar 4. Betty Boop lrange nonfriends 0 -1 1. Peter Pan 2. Batman
- 16. REDIS Pub/Sub.... a freaking AWESOME messaging queue
- 17. REDIS Pro’s: - Highly performant - Highly scalable - Interfaces available for many languages (Ruby: Ohm) Cons: - Eats your server’s memory - Manually sharding
- 18. MONGODB • http://mongodb.org • Document store • Writes to disk • Hight scalability • Easy sharding
- 19. MONGODB • Organizes data using collections • Stores documents in BSON format, a binary-encoded serialization of JSON
- 20. MONGODB use users u = { name: “ivan”, company: “Sysdivision”, age: “32” } users.insert(u) users.find(); - { "_id" : ObjectId("4d5ea4e79faeda067681e214"), "name" : "ivan", company: “Sysdivision”, "age" : "32" }
- 21. MONGODB use users u = { name: “ivan”, company: “Sysdivision”, age: “32” } users.insert(u) db.users.find({name: “ivan”}); - { "_id" : ObjectId("4d5ea4e79faeda067681e214"), "name" : "ivan", company: “Sysdivision”, "age" : "32" } db.users.find(); db.users.find().skip(20).limit(10);
- 22. MONGOID class User include Mongoid::Document field :name field :birthday, :type => date field :money, :type => float, :default => 0.0 embedded_in :company end class Company field :name embeds_one :user end
- 23. MONGOID company = Company.create(:name => ‘Sysdivision’) company.create_user( :name => ‘John Doe’, :birthday => ’12/7/1982’, : money => 0 ) User.find(:name => ‘John Doe’) User.where(:name => /.*ohn.*/)
- 24. MONGOID User.any_of( { :name => ‘John Doe’ }, { :money.gt 10 } ) User.any_in(:name => [‘John’, ‘Foo’, ‘Bar’]) User.where(:name => ‘John Doe’).and(:money.gt 10) Post.all_in(:tags => [‘ruby’, ‘code’, ‘mongo’])
- 25. MONGODB Pro’s: - Highly performant - Highly scalable - Easy sharding - Ultra powerful querying - Interfaces available for many languages (Ruby: mongo_mapper, mongoid)
- 27. QUESTIONS? Ivan Belmonte http://ibelmonte.com ivan@ivanhq.net