The advantage of the relational model Flexible, suits and data model Can integrate multiple applications via shared data store Standard within and between organizations Standard interface language SQL Ad-hoc queries, across and within ‘data aggregates’ Fast, reliable, concurrent, consistent
The problems of the relational model Much data is tabular but some is not inherently tabular Big data: volume variety velocity distributed
The features and goals of NoSQL database Does not use relational model(tables) Does not use sql language Designed to run on distributed servers Most are open-source Built for the modern web Schema-less Eventually consistent Goal To improve programmer productivity To handle larger data volumes and throughput
JSON and Graphs JSON: JavaScript object Notation Represents a(JavaScript) object and its properties An object consists of a set of attribute-value pairs, including arrays of objects Has a ‘tree’ structure Originally used for transmitting data between computers Now the storage format for document databases Graph: A data structure consisting of nodes/vertices and arcs/edges Nodes represent entities Arcs represent relationships May be directed or undirected In a graph database: Nodes and arcs can have properties and types The emphasis is on relationships
Types of NoSQL Key-value store: Berkeley DB Document databases MongoDB Big table HBase Graph database
Nosql-1 Key-value store: Berkeley DB Key = primary key Value = anything (number array image JSON)
Nosql-2 Document database: (JSON)MongoDB Like a key-value db, except that ‘value’(document) is ‘examinable’ by the db, so its contents can be queried and updated Document = object represented as JSON file Examples: MongoDB…
Nosql-3 Column family: Big Table, HBase Columns rather than rows are stored together on disk. Makes analysis by column faster
Aggregate-oriented databases Key-value, document store and column-family are ‘aggregate-oriented’ databases Pros: Entire aggregate of data is stored together Less need for transactions Efficient storage on clusters/ distributed databases Cons: Hard to analyse across subfields of aggregates
Nosql-4 Graph database Neo4J Examples: Neo4J infinite Graph Tao
Distributed data the CAP theorem Consistency everyone always sees the same data Availability system stays up when nodes fail Partition Tolerance: System stays up when network between nodes fail.
ACID versus BASE ACID: Atomic consistent isolated durable BASE basically Available soft state eventual consistency
Some notable NoSQL user Google – Big Table- column storage Facebook Tao Giraph- graph database Amazon- simpleDB- column storage Instagram-Cassandra LinkedIn- MongoDB
