NoSQL databases are non-relational databases
noSQL databases are non-relational databases. It means they do not share the same characteristics of RDBMS. They do not have foreign keys for example. The structure of the data is ordinarily much less structured than relational databases.
Often with Big Data there is not enough time to parse the data and convert it into a form suitable for Relational databases. This makes it faster to save the data in a noSQL database.
noSQL databases like Cassandra are designed to have multiple nodes for writing and are fast for writing. They are faster than relational databases which are also not that great for writing to from multiple locations.
There are several different kinds of noSQL databases:
Key Value Pair
Column
Document
Graph
Multi Model
Key Value noSQL databases store in key value pairs.
Includes the following dbs:
Dynamo, Redis , Riak, Tokyo Cabinet / Tyrant, Voldemort, Amazon SimpleDB, and Oracle BDB.
This is similar to the key value format except the values are stored in a document form using JSON or XML. This allows for the values to be queried more easily.
Includes the following dbs:
Mongodb , CouchDB, Riak, Amazon SimpleDB, and Lotus Notes.
The two components of Graph-type NoSQL databases are:
node
edge
Each node is an entity, for instance, the people in a social network application.
On the other hand, the edge is the relationship among entities.
Every relationship has its properties and is typically represented by a line.
An edge usually features a direction; for instance, an arrow indicates superiority in an organisation’s hierarchy.
Note large scale deployments can become significantly complicated due to many entity types and relationships.
Includes the following dbs:
Neo4J, Infinite Graph, and FlockDB
Multi Model handle numerous data models against one integrated backend.
Includes the following dbs:
OrientDB, ArangoDB, and MarkLogic Server, Redis
Page Author: JD
Back: Overview of Tech