Please enable JavaScript.
Coggle requires JavaScript to display documents.
Non-Relational Data Bases - Coggle Diagram
Non-Relational Data Bases
Aggregate Data Model
describes how we interact with the data in the database
Each NoSQL solution has a different model
document
aggregate orientation
collection of related objects wished to treat as a unit
makes a natural unit for replication and sharding
manipulate data through aggregate structures
column-family
key-value
graph
NoSQL Databases
Key/Value Stores
the key is a unique value
keep the data in memory/to disc
Caching
reducing disk I/O
Memcached
used in web applications
EHCache
used in Java applications
availability over consistency in distributed deployments
Amazon’s Dynamo
Basho Riak
high availability
operational simplicity
fault tolerance data replication
automatic data distribution across the cluster
Voldemort
the lowest latency
linear scalability
a big, distributed, persistent hash table
Apache Cassandra
Linear scalability
proven fault-tolerance
high availability without compromising performance
the highest throughput
Document Databases
treat a document as a whole
avoid splitting a document into its constituent name/value pairs
indexing on the basis of its primary identifier/its properties
open-source document databases
MongoDB
built for modern application developers and for the cloud era
general purpose, distributed database
CouchDB
span computing environment from globally distributed server-clusters to web browsers
Sorted Ordered Column-Oriented Stores
Features
avoidy consuming space when storing nulls
storing data effectively
units of data are sorted and ordered on the basis of the row-key
column-family
acts as a key for the columns it contains
the row-key acts as the key for the whole data set
Google’s Bigtable espouses such model
HBase column-family store by GB
can be manipulated using the MapReduce infrastructure
Graph Databases
small records with complex interconnections
nodes connected by edges
capturing any data consisting of complex relationships
product preferences
eligibility rules
social networks
Graph databases
FlockDB
for managing wide but shallow network graphs
is designed for rapid set operations
Neo4j
is designed for multi-hop graph traversal
Definition of NoSQL
a class of products and a collection of diverse (sometimes related), concepts about data storage and manipulation
operate without a schema, allowing to freely add fields