Adopting NoSQL Databases Using a
Quality Attribute Framework and Risks Analysis
Hilda Mackin, Gonzalo Perez and Charles C. Tappert
School of Computer Science and Information Systems, Pace University, 861 Bedford Road, Pleasantville, NY, USA
hd79424w@pace.edu, hdiaz19@hotmail.com,76gonzalo@gmail.com, ctappert@pace.edu
Keywords: Distributed Databases, NoSQL Databases, Relational Databases, Quality Attributes.
Abstract: NoSQL has emerged in recent years to provide increased scalability and performance, and organizations have
a problem in choosing between traditional SQL and NoSQL databases. This research gives software engineers
and architects a way to select a NoSQL database for a particular big data environment and domain. It proposes
a Quality Attribute Framework and Risk Analysis of NoSQL databases that can measure quality metrics
associated with availability and security, which are critical to choosing the right NoSQL database for a given
domain and to making better software development and design decisions. The framework will help IT
departments align perceived risks of NoSQL database adoption with actual risks, helping IT managers in their
database adoption and in the identification of risk factors that affect the new database technologies. The
framework developed here will be finalized through a qualitative analysis of risk vectors via surveys of top
IT leaders and IT companies.
1 INTRODUCTION
In the past, relational databases were used for all tasks
to be processed by a database because of their query
capabilities, and transaction management features.
Traditional Relational databases were designed
for different hardware and software times and are
facing challenges in meeting the performance and
scale requirements of Big Data (Grolinger et al.,
2013).
The vital factor for a change in data storage was
the need to support large volumes of data in
distributed environments. Two companies in
particular Google and Amazon have been influential
with their Big Table from Google and Dynamo from
Amazon papers.
One of the first solutions to increase the amount
of data that could be stored by a DBMs system was to
spread the data among several database servers or
clusters instead of just one server.
A second important factor for a change was that
distributed systems were not very efficient
performing transactions and join operations using a
Relational Database System (RDBMS).
Since Relational databases need to maintain strict
consistency using the transactional ACID model and
must be highly available a new family of NoSQL
clustered databases emerged.
As more companies’ considered adopting Big
Data Solutions, discussions about the most
appropriate NoSQL database for their use case,
application or environment originated.
NoSQL databases are highly scalable, have good
performance, are designed to store and process a
significant amount of unstructured data faster than
Relational Databases. So why to be cautious about
adopting a NoSQL database? Availability and
Security are major concerns for IT Enterprise
Infrastructures. There are perceived Security Risks
associated with the new database technologies
(Obijaju, 2015).
2 RELATED WORK AND
RESEARCH CONTRIBUTIONS
NoSQL data stores are seen as data processing
alternatives that can handle considerable volume of
data and provide better scalability, but attributes and
risks associated with these new technologies are not
well understood. Because of the large number and
diversity of NoSQL solutions, it is challenging to
97
Mackin H., Perez G. and Tappert C.
Adopting NoSQL Databases Using a Quality Attribute Framework and Risks Analysis.
DOI: 10.5220/0006227600970104
In Proceedings of the Fifth International Conference on Telecommunications and Remote Sensing (ICTRS 2016), pages 97-104
ISBN: 978-989-758-200-4
Copyright
c
2016 by SCITEPRESS Science and Technology Publications, Lda. All rights reserved
choose an appropriate NoSQL solution for a specific
task or Use Case. Some studies have identified
NoSQL challenges including the immense diversity
and inconsistency of terminologies, limited
documentation, sparse comparison and
benchmarking criteria, and lack of standardized query
languages (Grolinger et al., 2013).
Many researchers have focused on Performance
Evaluation but have not included other Software
Quality Attribute requirements (Lourenço et al.,
2015).
A previous study created a comparison of NoSQL
databases, identifying the software attributes that
would aid a software engineer’s decision process
(Lourenço et al., 2015). It identified several desirable
quality attributes to evaluate NoSQL databases:
Availability, Consistency, Durability,
Maintainability, Read Performance, Recovery Time,
Reliability, Robustness, Scalability, Stabilization
Time and Write Performance. It also selected some
popular NoSQL databases: Aerospike, Cassandra,
Couchbase, CouchDB, HBase, Mongo DB, and
Voldemort. Their study resulted in a summary table
(Table 1) to aid software engineers and architects in
their decision process when selecting a given NoSQL
database according to certain quality attributes.
Our research investigates, in a qualitative manner,
the attributes of an effective NoSQL Quality Attribute
Framework. This work will make two new
contributions to the state-of-the-art.
1) Provide a straightforward and coherent way for
IT Managers to understand NoSQL database quality
attributes and gain some insight on mitigation
strategies for current NoSQL databases risks.
2) Provide guidance to rationalize risks associated
with Security and Business Availability.
This paper improves and complements the study
(Lourenço et al., 2015).
1) Quality attributes are clearly defined (Section
4.2) and classified and evaluated by four types of
NoSQL database types: Key-Value, Document,
Columnar and Graph Database (Section 4.4).
2) This research adds Security attributes,
enhances Availability attributes and adds Additional
Quality Attributes to provide a complete set of quality
attributes. (Table 4).
The framework will be completed shortly through
a quality-attribute-focused survey based on NoSQL
database type (Key Value, Document, Columnar, and
Graph), where databases are compared with regards
to their suitability for quality attributes.
2.1 Research Approach
The goal of this research is to gain a thorough
understanding of NoSQL databases to clarify the risks
associated with these new technologies and to provide
a framework and recommendations to mitigate risks.
Two main research problems and a sub problem are
identified:
First Research Problem
What NoSQL databases Quality Attributes have
the most positive and negative impacts on Security,
and Business Availability risks?
What Software Quality Attributes are required to
evaluate and adopt NoSQL databases?
What is the perceived Business Availability and
Security Risks associated with these new database
technologies in the context of Big Data?
Second Research Problem
Are there any misalignments between actual and
perceived NoSQL database risks uncovered through
surveys and the risks perceived by experts and IT
professionals?
What are the technical and non-technical impacts
on the organization?
First Research Sub-problem
Are Relational and NoSQL databases totally
mutually exclusive?
How do applications intersect between Relational
and NoSQL databases?
If the technologies do not overlap, but intersect
choose between the databases?
3 NOSQL DATABASES
The NoSQL movement is a contemporary approach
to data persistence using novel storage methods
(Franks, 2012). NoSQL databases were built to deal
with the increasing amount of complex data, required
in some real-time applications, to address availability
over consistency, and to allow horizontal scalability,
using a distributed architecture and open source
(Bazar & Cosmin, 2014).
The common characteristics of NoSQL databases
are (Sadalage & Fowler, 2013):
Not using the relational model
Running well on clusters
Open source projects
Built for the 21st-century web estates
Schema-less.
Fifth International Conference on Telecommunications and Remote Sensing
98
3.1 Data Model
There are four accepted types of NoSQL databases:
Key Value Stores, Consist of keys and their
corresponding values, data is stored in a schema-less
way. This allows search millions of values in a
fraction of the time needed by conventional storage
(Franks, 2012).
Document Stores, Consist of a set of documents
possibly nested. Data can be structured in a schema-
less way or in the form of collections. Popular
examples are Couch Base, Mongo DB (Franks,
2012).
Column Stores or Extensible Record Stores,
Consist of tables which may have different schema
for each row having one huge extensible column
containing the data. Column stores do not declare null
fields as relational databases do (Franks, 2012).
Graph Stores, Consist of a set of graph nodes
linked together by edges (providing index-free
adjacency of nodes) (Barmpis & Kolovos, 2012).
Each type of NoSQL database is tailored for
storing a different and specific type of data (Barmpis
& Kolovos, 2012).
3.2 Advantages of NoSQL Databases
Most NoSQL data stores run on clusters. Relational
databases use ACID transactions to handle
consistency across the entire database. In contrast in
a cluster environment, NoSQL databases offer a
range of options for uniformity and distribution.
NoSQL databases operate without a schema,
allowing one to freely add fields to database records
without having to define any changes in structure
first. This is particularly useful when dealing with
non-uniform data and custom fields.
As a result, Relational databases are seen as just
one option for data storage. The result is that many
organizations will have a mix of data storage
technologies or Polyglot Persistence.
Besides handling data access with sizes and
performance that demand a cluster, the other
important reason for NoSQL technology is the
impedance mismatch problem. Big Data concerns
have created an opportunity for people to think in a
different way about their data storage needs. Some
development teams see that using a NoSQL database
can help their productivity by simplifying their
database access. This can be achieved even if they
have no reason to scale beyond a single machine,
improving the productivity of application
development by using a more suitable data interaction
style (Sadalage & Fowler, 2013).
4 WORK PLAN
The work plan will be composed of the following
sections:
1. Presenting a previous study’s comparison of
NoSQL engines and their quality attributes.
2. Defining a framework by enlarging the quality
attributes to include Security attributes,
enhancing Availability attributes and adding
Additional Quality Attributes.
3. Two additional NoSQL engines are added to
the study, illustrating two examples each of
the four types of NoSQL databases.
4. NoSQL databases proposed are categorized
into the four NoSQL database types.
5. The quality attributes of the two examples of
each type are consolidated to arrive at a direct
comparison of the four NoSQL database
types.
6. Finally, the set of attributes are enlarged to
include Security, Availability and Additional
Attributes, obtaining a framework to form the
basis for a subsequent qualitative study.
4.1 Previous NoSQL Database Study
A previous study created a comparison of NoSQL
databases, identifying the software attributes that
would aid a software engineer’s decision process
(Lourenço et al., 2015).
Their study resulted in Table 1 to aid software
engineers and architects in their decision process
when selecting a given NoSQL database according to
certain quality attributes.
The table uses a 5-point scale ranging from
“Great” to “Bad” to allow a direct comparison among
databases. For example, Cassandra is written based
on a performance-oriented approach, more than
Couch base. Worse grades were assigned when a
database was not an ideal pick, according to their
authors’ literature revision. This does not mean that
the database lacked the attribute entirely, but it was
not the best compared to the others databases
(Lourenço et al., 2015).
The quality attributes have the following
meanings:
Availability: Downtime was used as a primary
measure.
Consistency: Graded according to ACID
semantics consistency and how much consistency can
be fine-tuned.
Durability: Measured according to the use of
single or multi-version concurrency control schemes,
Adopting NoSQL Databases Using a
Quality Attribute Framework and Risks Analysis
99
the way the data are persisted to disk.
Maintainability: Measured for ease of setup and
use, accessibility of tools to interact with the
database.
Read Performance: Considered studies about the
fine tuning of each database.
Write Performance: Considered studies about the
fine tuning of each database.
Recovery Time: Related to availability, took
results from a previous study.
Reliability: Looked at synchronous propagation
modes
Robustness: Considered the tendency of databases
to have problems dealing with crashes or attacks.
Scalability: Looked at each database elasticity,
horizontal scaling, and ease of online scalability.
Stabilization Time: Related to availability.
Table 1: Quality attributes for popular NoSQL engines
(redrawn from (Lourenço et al., 2015)). Legend: “G” =
Great, “+” = Good, “A”= Average, “_” = Mediocre, “B” =
Bad, and “?” = Unkown/NA.
The study concluded that even though there was a
variety of other research and evaluations of NoSQL
technology, there was still not enough information to
verify how fit a NoSQL database is in a specific
scenario or system, making the following
recommendation for future work: “The development
of a framework for assessing most of these quality
attributes would greatly benefit the database adoption
of software engineers and architects” (Lourenço et al.,
2015).
4.2 Proposing a Framework
The framework proposed covers Availability and
Security attributes highly valued by NoSQL
databases users. Definitions of the quality attributes
included in the study are given below:
Availability: The percentage of time a system is
operating correctly. Is the data always accessible? Is
the data permanently available? In the context of the
CAP Theorem Availability was evaluated versus
Consistency and Availability versus Partition
Tolerance.
Consistency: The valid and reliable data that is
saved in every cluster node. Is the data the same in
every replication on every cluster node? In the
context of CAP Theorem Consistency is assessed
when all nodes see the same data at the same time.
Partitioning: Defined as the data divided in
smaller segments to be allocated in different data
store tables. Is Horizontal Partitioning or Vertical
Partitioning allowed?
Replication: Keeping a copy of the data in
different databases and servers. Is Replication
transparency allowed? Is Replication considered in
different layers? Does Master slave and master
replication have one instance or have different
instances?
Scalability: Related to horizontal scaling. Is
Scaling achieved by replicating the data
synchronously or asynchronously?
Shared Nothing: Are all replica nodes allowed to
continue working even if they are disconnected?
Recovery Time: Related to the time it takes for
several NoSQL systems to recover from a node
failure.
Stabilization Time: Related to the time it takes for
the system to stabilize when that node re-joins the
cluster.
Reliability: System’s probability of operating
without failures for a given period.
Robustness: Defined as the ability of the database
to cope with error during execution
Durability: Property that guarantees that a
transaction that has been saved/commit in the
database will be committed even in the event that the
system crashes
Maintainability: Does the NoSQL database
provide features for easy maintainability,
administration, management and operation?
Read Write Performance: Is the NoSQL database
more robust on Reading operations than writing
operations?
Security: Defined as the software and the set of
management tools that protect the database against
attacks, hackers, and viruses. Some properties that
need to be analysed in the study are:
Authentication: Related to password and user’s
login. What types of authentication does the NoSQL
database provides?
Authorization: Defined as a set of read and write
permissions request on tables, creation of users.
What administrative functions the NoSQL database
provides?
Fifth International Conference on Telecommunications and Remote Sensing
100
Encryption: Does the NoSQL database provide
mechanisms that allow encryption techniques that
preserve data confidentiality? If not what
mechanisms are used to enforce data confidentiality?
Three levels will be explored in the study: Data at
rest, Client to server communication, Server to Server
connection.
Auditing: Does the NoSQL database provide
mechanisms that allow writing to the database or
Audit Logs?
Data integrity: Does the NoSQL database provide
mechanisms that allow data integrity such as ACID
or eventually consistent BASE? Could the NoSQL
database achieve different levels of Data Integrity?
Confidentiality: Data Confidentiality. Does the
NoSQL database provide different mechanisms to
preserve data confidentiality?
Documentation: Does the NoSQL database
provide End User Documentation? What levels of
documentation does it provide?
Additional Quality Attributes Additional
attributes will be addressed in the study as part of the
contribution envisioned. These include attributes
related to the ease of developing when using NoSQL
databases:
Popularity Depending on the Type on NoSQL
database (Key Value, Document, Columnar or
Graph) some databases are more popular than others.
Different aspects need to be analysed to determine
what database is applicable to the business case and
from that infer the popularity.
Maturity: Considering Maturity of the API, time
in the market, and Enterprise adoption.
Query Possibilities: Does the NoSQL database
provide SQL like query possibilities?
Concurrency Control: Does the NoSQL database
provide features to manage Concurrency Control?
Does it provide Optimistic or Multi Version
concurrency control?
Conflict Resolution: Does the NoSQL database
provide mechanisms to manage Conflict Resolution?
4.3 Selecting NoSQL Databases
Considering four NoSQL database types (Key Value,
Document, Columnar, and Graph) the study selects
eight popular NoSQL databases used by enterprises,
two databases to represent each type. These database
types are now compared with regards to their
suitability for quality attributes.
The database selection was based on literature
research and data collected from preliminary
interviews: Voldemort, Redis, Mongo DB, Couch
DB, Cassandra, HBase, Neo4J, OrientDB.
4.4 NoSQL Database Types
We now categorize these eight popular NoSQL
engines into the four NoSQL database types, taking
five of the engines from Table 1 (Voldemort,
MongoDB, CouchDB, Cassandra, HBase) and
adding three new engines (Redis, Neo4J, and
OrientDB).
Table 2: Quality attributes for eight popular databases, two
of each type. Legend: “G” = Great, “+” = Good, “A” =
Average, - = Mediocre, “B” = Bad, and “?” =
Unknown/NA.
Two examples illustrate each of the four types, as
shown in Table 2.
The attributes for the engines from Table 1 come
from (Lourenço et al., 2015), the attributes for Redis
and Neo4J from (Redmond, 2012), and the attributes
for Orient Database from (Orient Database 2016)
The quality attributes of the two examples of each
type are consolidated to arrive at a direct comparison
of the four NoSQL database types (Table 3).
Averaging the attribute values for the two example
engines of each type was done conservatively by
leaning toward the lower rating on the 5-point scale.
4.5 Expert Opinions and Positions
Due to the diversity of NoSQL solutions and four
different types of NoSQL databases, making the
choice of the most appropriate data store for a given
use case scenario will be easier with the framework
proposed. Table 3 is now extended by adding the
security and additional quality attributes discussed
above to obtain Table 4.
Adopting NoSQL Databases Using a
Quality Attribute Framework and Risks Analysis
101
With the framework illustrated in Table 4, we will
reach out for expert opinions and positions through
interviews and surveys. A quality-attribute-focused
survey will be created, based on NoSQL database
type (Key Value, Document, Columnar, and Graph),
where databases are compared with regards to their
suitability for quality attributes. The anticipated
survey results should allow completion of blank
spaces on Table 4.
This research selected a methodology based on
finding qualitative measures and understanding
quality attributes of NoSQL databases by leveraging
the knowledge of NoSQL database specialists and
other early adopters users. It will involve:
Survey of enterprises and experts
Anonymity of Participants
Iterations
Controlled Feedback
Statistical Aggregation of Group Responses
Research Tasks Inputs and Outputs.
Table 3: Averaged quality attributes for the four NoSQL
database types. Legend: “G” = Great, “+” = Good, “A” =
Average, - = Mediocre, “B” = Bad, and “?” =
Unknown/NA.
4.6 Surveys
Example of the Survey with three brief sections is
listed below.
Please select the database used:
Voldemort(Key Value)
Redis (Key Value)
Mongo DB (Document)
Couch DB (Document)
Cassandra(Columnar)
HBase (Columnar)
Neo4J(Graph)
OrientDB (Graph)
Table 4: Averaged quality attributes for the four NoSQL
database types with the framework proposed. Legend: “G”
= Great, “+” = Good, “A” = Average, “-” = Mediocre, “B”
= Bad, and “?” = Unknown/NA.
4.6.1 Availability Survey
1. Is Horizontal Partitioning allowed? Yes, No
2. Is Vertical Partitioning allowed? Yes, No
3. Is Scaling achieved by replicating the data?
(Please Select)
Synchronously
Asynchronously
Both
4. Are all replica nodes allowed to continue
working even if they are disconnected? Yes, No
5. Is the NoSQL database more robust on Reading
operations? Yes, No
6. Is the NoSQL database more robust on writing
operations? Yes, No
7. Is the data permanently available? Yes, No
8. Is the data the same in every replication on
every cluster node? Yes, No
9. Does the NoSQL database provide
mechanisms that allow ACID data integrity? Yes, No
10. Does the NoSQL database provide
mechanisms that allow eventually consistent BASE?
Yes, No
Fifth International Conference on Telecommunications and Remote Sensing
102
4.6.2 Security Survey
1. Does the NoSQL database provide mechanisms
that allow encryption techniques? Yes, No
If “Yes” Please name the encryption technique
If “No” what mechanisms are used to enforce
data confidentiality?
2. What encryption level does the NoSQL database
provide? (Please Select)
Data at rest
Client to server communication
Server to Server connection
3. Does the NoSQL database provide mechanisms
that allow Auditing or Audit Logs? Yes, No
4. Does the NoSQL database provide mechanisms
that allow Authentication? Yes, No
If “Yes” please mention the Authentication
method
5. Does the NoSQL database provide mechanisms
that allow Authorization? Yes, No
If “Yes” please mention the Authorization
method
6. Is Replication on the database allowed? Yes, No
7. How often are backups tested? Daily, Weekly
8. How often is Disaster Recovery Infrastructure
tested? Daily, Weekly
4.6.3 Additional Quality Attributes Survey
1. What is the maturity of the Database - API in
the market according to your experience)?
(Please Select )
Mature (5- 10 years)
Growing (2 - 5 years)
Start Up (0 to 2 years)
2. In what Business Area the NoSQL Database is
used in your environment? (Please Select)
Medical
Financial
Retail
Social Media
Other (Please Describe)
3. How long have the database been used in the
Business Area selected previously? (Please
Select)
Popular (5- 10 years)
Growing (2 - 5 years)
Start Up (0 to 2 years)
Other (Please Describe)
4. Does the NoSQL database provide SQL like
query possibilities? Yes, No
If “Yes” please mention the SQL Query name
5. Does the NoSQL database provide features to
manage Concurrency Control? (Please Select)
Optimistic concurrency
Multi Version concurrency
Other
If “Other” please mention the Concurrency
Control
6. Is the NoSQL database coder friendly? Yes, No
7. Is the NoSQL database used as a caching
Layer? Yes, No
8. Is the NoSQL database used on Real Time
Analysis? Yes, No
9. Is the NoSQL database used on Analytics? Yes,
No
10. Does the NoSQL database provide End User
Documentation? Yes, No
The result of this methodology will be a quality
attribute framework and risks analysis of adopting
NoSQL databases, which will aid software engineers
and architects in their decision process when selecting
a NoSQL database according to their software
quality, attributes requirements.
5 CONCLUSION
There are a number of NoSQL data stores that can be
classified into four different types. However, there is
no Quality Software Framework that can help
managers decide which NoSQL databases are the
most appropriate for their Business Use Case.
The diversity of NoSQL data stores present
challenges to differentiate and get a perspective of
which databases is the most suited, establishing paths
and opportunities for future research.
Sophisticated Security and Privacy provisions
need to be explored. At the corporate level,
companies and institutions need to develop software
technology that offers Security features at the
minimum similar if not better than the ones used by
Relational Databases.
Considering a previous study’s comparison of
NoSQL databases and their quality attributes, the
contribution of this research includes Security
attributes, enhances Availability attributes and adds
Additional Quality Attributes to define a Quality
Attribute Evaluation and Risk Analysis of NoSQL
databases framework that will benefit the NoSQL
adoption in the long term.
The framework proposed will help IT
departments align perceived risks of NoSQL database
adoption with actual risks measuring quality metrics
associated with Availability and Security, which are
critical to choosing the right NoSQL database for a
given domain and to making better software
development and design decisions, giving software
Adopting NoSQL Databases Using a
Quality Attribute Framework and Risks Analysis
103
engineers and architects a better way to select a
NoSQL database for a particular big data
environment and domain.
REFERENCES
Barmpis Konstantinos, Kolovos Dimitrios S. 2012,
'Comparative Analysis of Data Persistence
Technologies', Proceedings of the 2012 Extreme
Modeling Workshop, New York.
Bazar Christina, Cosmin Sebastian 2014, 'The Transition
from RDBMS to No SQL. A comparative Analysis of
three popular No Relational solutions: Cassandra,
Mongo DB and CouchBase', Database Systems Journal,
University of Economic Studies, Bucharest, Romania,
Romania.
Catell Rick 2010, 'Scalable SQL and No SQL Data Stores',
ACM SIGMOD Record, ACM Digital Library, New
York.
Doshi Kshitij A.,Zhong Tao,Lu Zhongyan,Tang Xi 2013,
'Blending SQL and NewSQL Approaches, Reference
Architectures for Enterprise Big Data Challenges',
International Conference on Cyber-Enabled Distributed
Computing and Knowledge Discovery (CyberC).
Fowler Adam 2014, 'Why use a NoSQL Database, and why
not?', Adam's Blog.
Franks Bill 2012, Taming the Big Data Tidal Wave, finding
opportunities in Huge Data Streams with Advanced
Analytics, 1st edn, Wiley.
Grolinger Katarina, Higashino Wilson A,Tiwari
Abhinav,Capretz Miriam 2013, 'Data management in
cloud environments: No SQL and New SQL data
stores', Journal of Cloud Computing, Springer Open,
London.
Gualtieri Mike 2010, 'No SQL and Elastic Caching
Platforms Are Kissing Cousins', Forrester,
http://blogs.forrester.com/application_development/20
10/02/nosql.html.
Himmel Azua Maria 2012, Qualitative Analysis of Cloud
Computing Risks and Framework for the
Rationalization and Mitigation of Cloud Risks, Pace,
New York.
Intersimone David 2010, 'The end of SQL and relational
databases?', Computer World, February 2010.
Kroenke David M., ADJ, Database Processing, 13th edn.
Lakshman Avinash 2008, 'Cassandra A structured storage
system on a P2P Network', https://www.
facebook.com/notes/facebook-engineering/cassandra-
a-structured-storage-system-on-a-p2p-
network/24413138919/.
Lourenço João Ricardo, Cabral Bruno, Carreiro Paulo,
Vieira Marco, Bernardino Jorge 2015, 'Choosing the
right NoSQL database for the job: a quality attribute
evaluation', Springer Open Journal, no. Journal of Big
Data.
Nepa Suryal, Ranjan Rajiv, Kwang Kim, Choo Raymond
2015, 'Trustworthy Processing of Healthcare Big Data
in Hybrid Clouds', IEEE Cloud Computing.
Obijaju Mike 2015, 'No SQL NoSecurity Security issues
with No SQL Database', June 2015, https://
blogs.perficient.com/enterpriseinformation/2015/06/22
/nosql-nosecuity-security-issues-with-nosql-database/.
Orend Kai 2012, Analysis and Classification of No SQL
Databases and Evaluation of their ability to replace an
Object Relational Persistence Layer, University of
Munchen.
Orient Database 2016, 'Orient DB', http:// orientdb.com/
orientdb/.
Penchikala Srini 2011, 'Distributed Cache as a No SQL
Data Store?', InfoQ, http://www.infoq.com/news/
2011/11/ distributed-cache-nosql-data-sto.
Redmond Eric, WJR 2012, Seven Databases in Seven
Weeks, The Pragmatic Programmers.
Robin Hecht, SJ 2011, 'No SQL Evaluation', International
Conference on Cloud and Service Computing.
Robin Hecht 2011, 'No SQL Evaluation', International
Conference on Cloud and Service Computing, IEEE
Computer Society, Hong Kong.
Sadalage Pradmod J.,Fowler Martin 2013, NO SQL
Distilled. A brief guide to the emerging world of
Polyglot Persistence, 1st edn, Addison Wesley.
Stonebracker Michael, 'New SQL: An Alternative to
NoSQL and Old SQL for New OLTP Apps'.
Tudorica Bogdan George., Bucur Christian 2011, 'A
comparison between several NoSQL databases with
comments and notes', 2011 10th International
Conference RoEduNet, IEEE.
Wikipedia, 'Cassandra', Wikipedia, https://en.wikipedia.
org/wiki/Cassandra.
Fifth International Conference on Telecommunications and Remote Sensing
104