Big Data and Analytics by Samiya Khan

Author:Samiya Khan
Language: eng
Format: epub
ISBN: 9798885304887
Publisher: Notion Press
Published: 2021-11-15T00:00:00+00:00

7.3 CLASSIFICATION OF NOSQL DATABASES

The common thing between all NoSQL databases is that they don’t support SQL. The four categories of NoSQL databases are given below.

7.3.1 Key-Value Stores

In such databases, all the operations are performed using a key. For example, if you have a session database, then data retrieval from this table will be via SessionID, the key for the table.

•Databases:

Dynamo (Amazon)

•Dynamo is Amazon’s offering. They provide it in the form of Database-as-a-Service.

•MemBase

•CitrusLeaf

•Voldemort (LinkedIn)

•Riak

7.3.2 Big Data Clones

This class of NoSQL databases is completely based on the Google’s whitepaper that introduced BigTable concept. Databases that fall under this category include –

•HBase

It is an Apache project that is implemented in Java. We will discuss this database in detail later in the chapter.

•BigTable (Google)

This is the base project, which is based on the whitepaper published by Google.

•Hypertable

It is a C++ implementation of the Google’s whitepaper.

•Cassandra

Facebook developed this database system. The team that developed Cassandra was also a part of the development team of Amazon’s DynamoDB. Therefore, Cassandra is majorly a combination of BigTable and DynamoDB.

7.3.3 Document Databases

This class of databases is extremely useful for storing data in XML or JSON format and it includes the following –

•MongoDB

•CouchOne

•OrientDB

•TerraStore

7.3.4 Graph Databases

These databases are based on the graph concept. The best use-case for such databases is social media network that can be simulated in the form of a graph with users as vertices and connections between them as edges. The graph concept is an efficient method for managing social media data. Included solutions are –

•FlockDB

•Neo4J

•InfoGrid

•Sones

7.3.5 CAP Theorem

The decision of which database to use from the given list is majorly dependent on your performance and system requirements. CAP theorem is a standard approach that is used for assistance in making such decisions. The CAP theorem uses three parameters namely, consistency, availability and partition tolerance and states that it is possible to get only two of these characteristics in a NoSQL solution. Therefore, the decision of which database to use must be purely based on which characteristics are important for you. The three characteristics are as follows –

1.Consistency

The commits performed in a database are atomic across the whole database.

2.Availability

The database is available and accessible at all times.

3.Partition Tolerance

The system responds correctly in all conditions except when there is a total network failure.

4.The distribution of databases across these characteristics is illustrated in Fig. 7.1.

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.