MySQL vs MongoDB

MySQL		MongoDB
	VS
http://www.mysql.com/		http://mongodb.org

MySQL is a relational database with fixed schema

• Internal structure proprietary to the storage engine and exposed through an SQL interpreter
• Best-practice to normalize to Third Normal Form or better

MongoDB is a non-relational NoSQL schemaless database

• Internal BSON structure (binary JSON-like format) exposed through a JavaScript interface
• Best-practice to de-normalize data

GUI Administration Tools for PHP

MySQL

• phpMyAdmin 30mb, hundreds of files, lots of configuration settings

MongoDB

• phpMoAdmin 90kb, single self-configuring file, Fast AJAX interface

Leadership

MySQL

• Oracle CEO: Larry Elison

MongoDB

• 10gen CEO: Dwight Merriman (cofounder of DoubleClick and many other successful startups)

Scaling

MySQL

• Typical: replication and clustering
• Other options: sharding and federated databases

MongoDB

• Typical: replication and sharding

Index Systems

MySQL

• B-Tree (default)
• Hash (Memory tables and NDB only)

MongoDB

• B-Tree (default)
• Geospatial 2D

Resource Utilization

MySQL

• Disk Space, MyISAM: Marginally larger than the data plus indexes
• Disk Space, InnoDB: Somewhat larger than the data plus indexes
• RAM: Allocates as much RAM as available; even more with memory tables. Improper MySQL configuration can leave the OS without any available RAM.
• CPU: Allocates a reasonable amount of CPU, relative to the task

MongoDB

• Disk Space: between 2x and 10x data size, minimum active-database size of 50mb
• RAM: Allocates a reasonable amount of RAM
• CPU: Marginal, relative to the task

Terminology

• MySQL table = Mongo collection
• MySQL row or record = Mongo document or object
• MySQL column = Mongo node of a document object

Related Videos

• Full-HD video of 10gen/MongoDB CTO Eliot Horowitz presenting MongoDB at Database Month
• HD-video of 10gen/MongoDB CEO Dwight Merriman presenting MongoDB to the NYC MySQL Group at Sun Microsystems
• From MySQL to MongoDB at Sluggy.com at the MongoNYC conference
• From MySQL to MongoDB at Wordnik at the MongoNYC conference