You are on page 1of 20

BASE analysis of NoSQL database

Deka Ganesh
Regional Vocational
Training Institute for
Women, Tura, Meghalaya
Ministry of Skill
Development and
Directorate General of
Training, Government of

NoSQL databases are designed to address

performance and scalability requirements of
web based application which cannot be
addressed by traditional relational databases
The relational databases strongly follow the
ACID (Atomicity, Consistency, Isolation, and
Durability) properties while the NoSQL
databases follow BASE (Basically Available, Soft
State, Eventual consistency) principles. This
survey paper is an analytical study on BASE
features of some of NoSQL databases.
The following three are the major
contributors of the NoSQL
Engage your Audience
BigTable (Google) started in 2004
by Google.
. Dynamo: Amazons Highly
Available Key-value Store.
CAP theorem
The common motivations of NoSQL design are:

Easier deployment.
Large scale data
Meeting the scalability and
Can be used as a Caching layer
for storing the transaction data
Key features of NoSQL are:
Horizontal Scaling/Scaling Out for
simple operations by adding more
machines to a pool of resources,
whereas, Vertical Scaling/Scaling
Up means adding more CPU, RAM,
etc. to existing machines.
The RDBMS can Scale Up, but not
Scale Out, hence suitable for
strong Consistency and Availability.
In most of NoSQL data is partitioned
and replicated across multiple
nodes. Inherently, most of them use
either Googles MapReduce or
Hadoop Distributed File System or
Hadoop MapReduce for data
NoSQL characteristics
Shared Nothing architecture
Asynchronous Replication
BASE instead of ACID
NoSQL categories

NoSQL databases are broadly

classified into following 2
Aggregate Oriented
Non-aggregate Oriented.
The Pros and Cons of Key/Value Databases are:
Pros Cons

Very fast Many data structures

Very scalable (objects), hence
Simple model. cannot be easily
modeled as keyvalue
Able to distribute
BASE Analysis

Basically available
The database system always seems to work
Soft state
By using soft state, the resources/functionality
which has been loaded during the initial load can be
reused for the subsequent requests of the service
Eventually consistent
The updates propagate, when there are no updates
over a certain period of time defied for a particular
application. Hence, the systems where BASE is the
key requirement for reliability, the potential of the
data changes essentially slows down. A lot of d
Benchmarking of NoSQL
Journal Critique

The article BASE analysis of NoSQL database by Deka

Ganesh Chandra seek to provide information and
analysis of NoSQL based on BASE (Basically Available,
Soft State, Eventual consistency). The research problem
is not clearly stated on the introduction but from further
reading, it indicate that the main goal of the article is to
provide a level of information and benefits of using
Nosql in contrast to Ralational database. The abstract
and the introduction focuses on the contrast between
SQL using Atomicity, Consistency, Isolation, and
Durability (ACID) properties and Nosqls Basically
Available, Soft State, Eventual consistency(BASE).
Journal Critique Continued
The author starts by giving the main contrast of NoSQL to SQL.
Indicated in the abstract that SQLs ACID versus NoSQLs BASE.
In the introduction, the common motivation on Nosql design are
indicated. Key features such as Horizontal Scaling/Scaling Out
for simple operations is also stressed out. The systems
discussed in this study are more representative rather than
comprehensive. This paper is organized into 7 sections. Section
2 is about the characteristics of NoSQL databases. Section 3 is
about categories of NoSQL databases. While Section 4 is BASE
analysis of NoSQL. Section 5 is a Discussion about the BASE
features of various NoSQL. Section 6 is about the benchmarking
of Cloud services with a focus on NoSQL. Section 7 is
Observations and recommendations for picking up NoSQL for a
particular application. Section 8 concludes the paper.
Journal Critique Continued

In section 2, contrast in Schema between

Nosql and SQL is emphasized. Nosql, using
schema-less approach while SQL using strict
schema approach. Characteristics of Nosql
is also discussed in some details. Schema-
less, Shared Nothing architecture, Elasticity,
Sharding and Asynchronous Replication are
provided with some information.
Journal Critique Continued

Section 3 discussed the classification of NoSQL under being

aggregate oriented and non-aggregate oriented. Examples of
these classification is also mentioned, such as Amazon S3
(Dynamo), Voldemort, Scalaris, Cassandra and MongoDB for
Aggregate oriented part. Neo4J is the given example for non-
aggregate oriented part of NoSQL. Difference between the 2
is also given, The basic difference between the Aggregate
oriented database and Non-aggregate databases is that
Aggregate Oriented database splits relations. Is mentioned.
Common features such as schema-less for both classification
is also mentioned.

Journal Critique Continued

Section 4, provides a much detailed BASE analysis for

Nosql. Discussion on what type os Nosql is suited for
different scenarios are also enumerated. . CouchDB is
a popular database for integrating the web and
mobile applications., Google is using Single Master in
Google file system (GFS). Consistency vs. Latency
issues are also briefly discussed citing examples on
the effects of latency on e-commerce sites, thus
putting emphasis on latency causing troubles on
these named sites. Nosql provides solutions on
latency issues which makes Nosql better suited for e-
commerce sites.
Journal Critique Continued

Overall the article is straightforward and

direct in providing comparison and contrast
between sql and Nosql. As the abstract
clearly state, the article provides
performance analysis as well for some of the
Nosql systems available. Overall, it is a
good, well-written article with lots of
information about NoSql, its strengths in
relation to the purpose of the system being
Presented by:
Christian F Ramos