International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 07 Issue: 04 | Apr 2020 www.irjet.net p-ISSN: 2395-0072
© 2020, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 5734
Review Paper on Web Frameworks, Databases and Web Stacks
Jyoti Shetty
1
, Deepika Dash
2
, Akshaya Kumar Joish
3
, Guruprasad C
4
1-2
Professor,
3-4
UG Student
1-4
Department of Computer Science and Engineering
1-4
R V College of Engineering, Bangalore India
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - In today's world, online applications have
become an important component of the industry. Businesses
can now grow and become easier by using the web apps, and
achieve their goals much more quickly. The development of a
web application involves identifying product requirements,
designing, coding, and testing using frameworks and
technologies. Frameworks are libraries and classes that give a
rich collection of functionalities. Applications are built on top
of frameworks so that the fundamental requirements in
developing a web application are already setup. There are
various frameworks available for front-end, and back-end
development supporting different programming languages.
Database is where the web application stores its data for use
by back-end frameworks. The most important part of a web
development lifecycle is choosing the right combination of
front-end framework, back-end framework, and database
environment. Web stack is a combination of frameworks,
database environment, and an optional operating system
environment. This paper discusses a few of the important
front-end frameworks, back-end frameworks, databases, and
web stacks.
Key Words: Microservices, Framework, Database, Web
Stack.
1. INTRODUCTION
In this world of internet, where there are numerous
technologies that are available for web application
development, choosing the right architecture for application
development is a very difficult task. One such architecture is
the Microservices based architecture. In this approach, the
various collections of services are developed as loosely
coupled, highly maintainable, and independently deployable
services. From a web application perspective, the application
can be broadly divided into two independent services viz.
Frontend service and Backend service. These two can
interact with each other through API calls. The frontend of
the application handles the user interaction while the
backend of the application handles the business logic. Having
a separate service handling the frontend and backend
independently gives a lot of advantages like fast
development, easy up-gradation, and high maintainability.
But this also comes with its problems which is the risk of
running two separate instances for one application. This
hampers the continuous integration/continuous deployment
pipeline. This paper discusses the various frameworks,
languages, and solution stacks that are available for
developing a microservices-based web application.
The rest of the paper is organized as follows: Section 2
focuses on the frontend frameworks and languages, Section
3 is about backend frameworks and languages, Section 4 is
about databases followed by Section 5 which describes
various web stacks finally Section 6 concludes this review
paper.
2. FRONT-END FRAMEWORKS AND LANGUAGES
The component of an app or website that users interact with
is called Front End or Client Side of the application. It
includes everything that users come across directly including
images, buttons, text colours, graphs, tables, etc. HTML and
CSS are used for the design and styling while JavaScript is
used for validations. The behaviour, design, content, and
structure of everything visible on the screen when
mobile/web applications are opened, is developed by Front
End application developers. Performance and
responsiveness are the two main objectives of the front end
development. The developers must make sure that the
website is responsive i.e. it displays correctly on devices of
all sizes. No part of the app/website should behave
irregularly irrespective of the screen size.
2.1 AngularJS
AngularJS is a JavaScript framework for developing dynamic
single page web applications. It is developed and maintained
by Google. It uses HTML as a template language and extends
HTML's syntax to express the application's components
clearly and concisely. AngularJS's dependency injection and
data binding eliminate much of the code which might
otherwise be needed to be written and it all happens within
the browser, making it apt with any server technology.
2.2 ReactJS
ReactJS is a JavaScript library for developing dynamic User
Interfaces. It's developed and maintained by Facebook. React
is an efficient, declarative and versatile JavaScript library for
building user interfaces. Complex UIs can be composed from
small, isolated and reusable pieces of code called
"components". ReactJS uses JSX to simplify writing HTML. JSX
is a pre-processor that adds XML syntax to JavaScript.
International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 07 Issue: 04 | Apr 2020 www.irjet.net p-ISSN: 2395-0072
© 2020, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 5735
2.3 VueJS
Vue is a progressive framework for building user interfaces
and single-page web applications. Vue is designed to be an
incrementally adaptable architecture that focuses on
declarative rendering and component composition. VueJS is
created and maintained by Evan You. Vue extends HTML
with HTML attributes called "directives".
Table -1: Comparison of AngularJS and ReactJS
Comparison
Parameter
AngularJS
ReactJS
Baseline
Highly performing
JavaScript
framework
Excellent JavaScript
library
Why Choose
If using
TypeScript coding
-is-
Founders
Maintained by
Google
Maintained by
Facebook
Learning Curve
Steep learning
curve
Fairly simply
compared to
Angular
Model
Based on Model
View Controller
Based on Virtual
DOM
Written in
TypeScript
JavaScript
Community Support
Very good
community
support
Very good
community support
Popularity
Used by ~20%
developers
Used by ~75%
developers
3. BACK-END FRAMEWORKS AND LANGUAGES
Web creation at the Backend consists of countless activities.
For example, protecting APIs (Application Programming
Interfaces) against external attacks, authenticating users,
enabling seamless interaction with databases, and handling
user requests to collect and present the required
information, etc. The backend frameworks enable all of these
activities to developers simple and trouble-free.
These frameworks are evaluated by their programming
methods, languages they support, and interfaces.
Additionally, frameworks that provide built-in tools and
templates that help developers in tracking various web
development tasks quickly. Directly put, an advanced
backend architecture increases the pace of development,
thereby saving development time. These frameworks should
allow developers to create interoperable applications that
can scale up the workload. A few of the important
frameworks are discussed in the following section.
3.1 NodeJS
Node.js is an asynchronous event-driven JavaScript runtime
designed to build scalable network applications. It supports
the handling of many connections concurrently. Call back is
triggered for each connection and it sleeps if there is no
work to be done. This behavior unlike too many other
common concurrency models, in which operating system
threads are used. Moreover, the users of Node.js need not
worry about deadlocks, since there are no locks. Since
almost no function performs an I/O in Node.js, the process
never blocks. These properties make Node.js scalable.
JavaScript language is used to write applications in Node.js
and run with Node.js runtime environment. The package
ecosystem of Node.js called npm (node package manager)
has a large number of open-source libraries. Node.js uses
-fast highly optimized V8 execution
engine in JIT (Just in Time) compilation fashion to execute JS
code by transforming them into machine language and
optimizes through complicated methods such as code in-
lining, copy emission, etc.
3.2 Django
Django is a web framework based on python programming
language. It is ridiculously fast and was designed to help
developers transform concept to application as quickly as
possible. It takes care of much of the struggle of web
application development thereby allowing developers to
focus on writing actual business logic of the app without
needing to reinvent the wheel. It is incredibly versatile.
Django takes care of many necessary tasks of web
development such as user authentication, site maps, content
administration, RSS (Really Simple Syndication) feeds, etc.
Famous web applications such as Instagram, Udemy, Mozilla
and Pinterest use Django. Django follows a Model-View-
Controller (MVC) architecture. The Model is the logical data
structure that acts as a mediator between the website
interface and the database. The View component contains
the user interface logic. The controller is the main control
component and acts as the middleman that binds the view
and model together, meaning it enables the data passing
from model to view.
3.3 Spring Boot
Spring boot sticks well with microservices but not
encouraged for large monolithic applications. The artifacts
can be deployed directly into Docker containers. Spring boot
eases the development of web applications. In standard code
and setup related to web advancement, a cutting-edge web
development model that smooths out the improvement of
server-side UI applications, REST APIs, and bidirectional,
event--
ready features such as health status monitoring, metrics
data, and tracing helps developers to monitor and get insight
into the application. It supports many industry-standard
authentication protocols like SAML (Security Assertion
Markup Language), OAuth (Open Authorization), and LDAP
(Lightweight Directory Access Protocol). It supports various
relational and non-relational databases, cloud-based data
services, and map-reduce frameworks such as Spark, Flume,
etc. Spring supports Java, Kotlin, and Groovy. The
development time is considerably more compared to other
frameworks because of the programming language and
configurations. But this has a use case where there is
International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 07 Issue: 04 | Apr 2020 www.irjet.net p-ISSN: 2395-0072
© 2020, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 5736
concern about enterprise acceptability, standard,
integration, and security.
Table -2: Comparison of Node.js and Django
Comparison
Parameter
Node.js
Django
Costs
Open Source
Open Source
Scalability
More scalable
comparatively
Less scalable
Architecture
Event-driven
programming
Model-
Template-View
Security
Developers need
to need to
explicitly make
sure the security
provided
Provides great
security and
developers
need not worry
about basic
security
concerns
Community
It has a very
active
community that
helps
programmers to
improve
themselves and
in improving the
framework
It has few
communities
for support
since it is not as
famous as
Node.js.
Leading
Many countries
are using Node.js
and it is ahead
comparatively
It is new and
behind Node.js
in usage
Performance
Its performance
is better
Its performance
is good
Complexity
It is less complex
It is more
complex than
Node.js
4. DATABASES
When it comes to choosing a database, the main choices are
to choose between relational (SQL) and non-relational
(NoSQL) databases. Although both databases are viable
alternatives, some key differences still remain between the
two that users need to bear in mind when making a decision.
Each database is built for a specific type of use case and it is
very important to choose the right database based on
application needs. The relational databases are suitable for
small scale data and to store OLTP data. Non-relational
databases are preferred when the data is unstructured and
the developer cannot stick to a fixed schema.
4.1 PostgreSQL
PostgreSQL is ideal for OLTP (Online Transaction
Processing) type of workloads because of its compliance
with ACID properties. It is a good choice for its OLTP
capabilities. PostgreSQL can act as a highly capable analytical
database. Mathematical software such as MATLAB can be
easily integrated with PostgreSQL. PostgreSQL is highly
scalable and it meets the requirement of modern websites
that require thousands or even hundreds of thousands of
requests per second to serve customers. PostgreSQL
integrates well with all modern web frameworks such as
Django, Hibernate, Node.js, PHP etc. It has excellent
replication capabilities which helps in easily scaling out to as
many database servers as required. PostgreSQL does not
limit itself to just a relational database. It can serve as a
NoSQL style data store. In applications that need to leverage
both relational and document-oriented databases
PostgreSQL is a good choice as the developer can have both
in a single product.
4.2 MongoDB
MongoDB is a NoSQL database. It is document-oriented,
rather than using tables and rows as within the traditional
SQL databases, MongoDB uses collections and documents.
Documents contain key-value pairs which are the
fundamental unit of data in MongoDB, it is equivalent to
records in relational databases. Documents of similar type
allows easy representation and storage of hierarchical
relationships in the form of arrays, and other complex data
structures. MongoDB is very scalable with replication
capabilities. BSON is the format used to represent JSON
documents in MongoDB. BSON provides additional data
types such as Date and binary which are not supported in
JSON and also provide ordered fields in order for it to be
efficient for encoding and decoding within different
languages. MongoDB finds it use case in big data and real-
time applications running at multiple different locations.
While it resolves network partitions by maintaining
consistency, while compromising on availability.
Table -3: Comparison of PostgreSQL and MongoDB
Comparison
Parameter
PostgreSQL
MongoDB
Architecture
Monolithic
Distributed
Scaling
No built-in
mechanism to
horizontally
scale by
partitioning
Sharding is used
horizontally scale-
out database.
Commodity
hardware is used
for this purpose
Availability
Does not have
integrated
mechanisms for
handling
automatic
failover and
recovery
between
replicas
Maintains enough
number of
replicas to
achieve
availability and
has automatic
failover handling
mechanisms
JSON Data handling
Drivers lack the
ability to
accurately
convert JSON to
useful datatype
used in
application
BSON and its
drivers support
more complex
and advanced
data types which
are not supported
by JSON
International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 07 Issue: 04 | Apr 2020 www.irjet.net p-ISSN: 2395-0072
© 2020, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 5737
Data Governance
Does not have
any in-built
mechanism for
JSON schema
validation
Supports schema
validation
through the IETF
standard, thereby
enforcing schema
structure on
database
collections
Schema
Strongly typed
Flexible
Transaction
ACID
Multi-document
ACID
Transactions with
snapshot isolation
Consistency
Achieves
Immediate
Consistency
Supports both
Immediate
Consistency and
Eventual
Consistency
which can be
individually
decided for each
write operation
5. WEB STACK
A web stack also known as web application stack is a
collection of software applications needed for web
development, specifically for developing web applications
and implementing websites. A web stack is a type of solution
stack, which is a collection of software for performing
specific tasks. Web stacks are critical components for web
applications as well as websites. At a minimum, a web stack
should contain an operating system, database software,
programming language, and the webserver.
5.1 LAMP/LEMP
Linux, Apache/Nginx, MySQL, PHP
This is a popular web stack which involves an operating
system, Linux, a web server, either Apache or Nginx, a
relational database management system, MySQL and a
scripting language well suited for web development, PHP.
The scripting language can also be Perl or Python but, PHP is
the popular choice. This stack is well suited for developing
dynamic web sites and web applications. This is a layered
structure and each layer builds on the previous layer. Linux
is the base layer with the operating system. Apache/Nginx is
the web demon that sits on the OS. The DB layer is used to
store the data and the scripting language is used to drive and
display the data and enable user interfaces.
5.2 MEAN/MERN/MEVN
Mongo, ExpressJS, AngularJS/ReactJS/VueJS, NodeJS
This is a highly popular JavaScript based software stack used
to build dynamic websites and web applications. The major
advantage of this stack is that both the client side and server-
side environments can be developed in one language.
MongoDB is a NoSQL database and is used for data storage.
ExpressJS is a JavaScript library used as a reverse proxy for
an application server like Apache or Nginx. Angular, React
and Vue are popular JavaScript frameworks for developing
the client-side application framework. NodeJS is a JavaScript
runtime environment that can execute JavaScript outside a
web browser. This is where the server-side scripts will
execute.
5.3 SPRING STACK
Using Java Spring as the base calls for the Spring Stack. This
stack is used to build dynamic websites and web
applications. Spring Boot provides various starter packs to
start development very quickly without doing any
configurations for application servers. It is very easy to build
REST services using it. Database connectivity can be greatly
simplified by using JPA and Hibernate. The frontend
framework can be chosen from a wide variety of choices like
ReactJS, AngularJS, JSP etc. Spring Boot also provides an
embedded application server when required.
6. CONCLUSION
Every web framework, database and web stack has its own
set of advantages and disadvantages. The choosing of web
stack should be done based on few critical factors like time
needed to take the product to market, long term scalability
proficiency with the technologies etc. Database has to be
chosen based on the type of data the application is dealing
with such as structured, unstructured or semi-structured
and the consistency, availability and partition tolerance
required by the application. All these factors are important
to be considered while building modern web applications.
REFERENCES
[1] Angular.io. 2020. Angular. Accessed on 29 February 2020
[online] Available at: <https://angular.io/docs>
[2] Reactjs.org. 2020. React Accessed on 06 March 2020
[online] Available at:
<https://reactjs.org/tutorial/tutorial.html>
[3] P. Di Francesco, "Architecting Microservices," 2017 IEEE
International Conference on Software Architecture
Workshops (ICSAW), Gothenburg, 2017, pp. 224-229.
Integration with Java Artificial Intelligence Weka
CONFERENCE, GABROVO
[5] Dunka, Bakwa & Emmanuel, Edim & Oyerinde, Yinka.
(2018). Simplifying Web Application Development Using-
Mean Stack Technologies. 04.
International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 07 Issue: 04 | Apr 2020 www.irjet.net p-ISSN: 2395-0072
© 2020, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 5738
[6] Mongodb.com. 2020. MongoDB. Accessed on 07 March
2020 [online] Available at:
<https://www.mongodb.com/compare/mongodb-
postgresql>
[7] Nodejs.org. 2020. Nodejs. Accessed on 28 April 2020
[online] Available at: <https://nodejs.org/>
[8] Docs.djangoproject.com. 2020. Django. Accessed on 26
April 2020 [online] Available at:
< https://docs.djangoproject.com/en/3.0/>
[9] www.techopedia.com. Web Stack. Accessed on 01April
2020 [online] Available at: <www.techopedia.com>
