Luận văn electronic medical record emr database and big data using microsoft azure projec

In place of buying, owning, and maintaining data centers and physical servers, we can get into technology services, such as computing power, sLorage, and databascs from cloud service pr

Electronic Medical Record (EMR) Database

And Big Data Using Microsoft Azure Project

Student’s name

ĐỒ THỊ TÚ UYÊN

Hanoi - Year 2020

VIETNAM NATIONAL UNIVERSITY, HANOI

INTERNATIONAL SCHOOL

œe EH] cam

GRADUATION PROJECT

PROJECT NAME Electronic Medical Record (EMR) Database And Big Data Using Microsoft Azure Project

SUPERVISOR: Assoc Prof, Dr Nguyén ‘Thanh ‘Tung STUDENT: Đề Thị Tủ Uyên

STUDENTID: 16071317 COHORT: MIS2016A

MAJOR: Management Information System

Hanoi - Year 2020

FORM 3: LETTER OF DECLARATION

Thereby declare thal the Graduation Project “Electronic Medical Record (EMR) Database And Big Data Using Microsoft Azure” is the results of my own research and has never been published in any work of others During the implementation process of

this project, I have seriously taken research ethics, all findings of this project are results

af my own research and surveys: all references in this project are clearly cited according

FORM 4: ACKNOWLEDGEMEN

First of all, I would like to express my sincere thanks to all teachers in International School

— Vietnam National University far teaching and imparting valuable knowledge to me

during the four years of studying and training at the school I want to express my deep

gratitude to the instructor - Assoc Prof Dir Nguyen Thanh Tung {and his assistants MR

Vu Khanh Ioan), and my old teacher Mr-Phat, who enthusiastically guided me to implement this thesis as well as creatmg favorable conditions for me to have real contact, helping me gain more knowledge, practical and future job requirements

With my limited knowledge and research time, my thesis cannot avoid shortcomings I look forward to receiving comments and suggestions from leachers Ii will be a valuable

experience for me to improve myself later

Sincerely thank!

FORM 5: TABLE OF CONTENT

CHAPTER 1: OVERVIEW OF BIG DATA ON THE CLOUD COMPUTER PLATFORM

TL Cloud Computing

TA Tntroduction about Cloud Computing

T2 Components and structure of cloud computing,

13 The Services of Cloud Computing

13.1 SaaS: Sofware as a Service

13.2 PaaS: Plalform as a Service

1.3.3 TaaS: Infrastructure as a Service

14 Models of Cloud Computing

15.2 Some applications of Cloud Computing

16 Typical cloud applications

L6.1 Google App Engine

16.3 Amazon Web Services

162 Windows Azwre

17 The situation of apply Cloud Computing in the world and Vietnam

IL Big Data - Overview And Its Impact On Health Care co 24

ILI The definition of Big Data and What is Big Data in Health Care System and Why we

TLL 2 What Is Big Data In Health Care System? And Why We Need It? 24

ILS The application and important of Big Data in Ilealth Care System 3⁄4

IL6 What is NoSQL System? Why did Ï choose Ì, oousueee sao reo oe 37

ID Overview aboutCassandra CosmosDD oeneeeoeeree seo AO

TILL Introduction, Definition, and features nvsesnmenesnemennnen nannies muse a AO TIL2, Strength and Weakness snensinmneisnnnnnennaenesennnee sstnines sees AZ CHAPTER 2: APPLICATION AND RESULTS ACIHBVEĐ sec o đ8

1 Introduction about the tool Microsoft AZure je suonnnenmnenn sinnerman AB

IL, Application in the project - health care systeim soi neeeoeieerr semen ĐT TLL About the data of the topic csse:smmneininmnnnnenienmannnnnatnaies ssenans Sh

12 8eup Biy Data

113, Demo the resul

I — Conclusion

ILL Assessment Of Project

T112 Contributions of the project

1113 The Limitations of the project

References

63

65

+ 6S

65

67 69

FORM 6: TABLE OF NOTATIONS AND ABBREVIATIONS

Version

Interface

FORM 7: LIST OF TABLE AND FIGURE

1-thrcc types of cloud computing

2-comparison between SQL and NoSQL

3-the regions serving of Microsofi Azure

4- data filed

S-data ficld 2

7-create Big Data l

8-create Bip Data 2

9-create Big Data 3

10-create Big Data 5

11-ereate Dig Data 6 cu nh he

12-create Big Data 7 co né Han

13-creale Big Data 8

l4-create Big Data Ø eects vier vanes

1S-create Big Data 10

16-Create Table in Big Data 1

17-Create Table in Big Data 2

ABSTRACT

Today, for hospitals, the good and officicnt management of hospital data as well as paticnL

data is one of the top priority problems and is constantly causing difficulties for them

To manage thal data, hospilals initially have to invest and calculate many types of costs

such as hardware, software, network costs, administrator costs, maintenance costs ,

ropaired, Besides, they also have to calculate the ability to expand and upgrade equipment,

Data security must be controlled as well as high availability

From such a typical problem, we see that if there is a trusted place to help hospitals manage

that data well, they will no longer care about infrastructure and technology but only focus

Mainly, their medical examination and treatment will bring them more efficiency and profit,

as well as more convenierice for patients

‘The terms "big data” and "cloud computing" came from one of these circumstances The term is also derived from the idea of putting everything such as data, software, calculalions, etc on the Internet We will no longer see computers, servers of hospitals alone to store

dala and software, but only scveral "virtual servers" gathered on the network "Virtual

servers" will provide services that make it easier to manage data; they will only pay for

their service usage, without having to invest heavily in infrastructure as well as paying

much attention to technology

This trend not only helps hospitals, but also helps many other small and medium-sized

enterprises and businesses that do not have sufficient network infrastructure, servers to

store and manage data well

Ilowever, due to time and knowledge constraints, in this report I will only delve into Big Data in the field of health care on Cloud computing platform Simultaneously, demo

application of Microsoft Azure sofware to manage patient information

CHAPTER 1: OVERVIEW GF BIG DATA ON THE CLOUD COMPUTER PLATFORM

1 Cloud Computing

1.1 Introduction about Cloud Computing

Cloud computing 1s the most significant thing in the information technology industry,

almost all of the biggest critics agreeing cloud computing is one of the most important paradigm changes in the early 21st century But that's all and where does it come from?

And how docs cloud compuling history take shape?

According to Amazon, “Cloud computing 1s an on-demand distribution of IT resources

aver the Internet with a usage-based payment policy In place of buying, owning, and

maintaining data centers and physical servers, we can get into technology services, such

as computing power, sLorage, and databascs from cloud service providers like Amazon

Web Services (AWS)” [1]

Cloud computing onginated in the 1950s when the mainframe was barn ‘The cost of

these mainframe devices is so large that organizations cannot afford to buy them

economically ‘This is the time when the idea of providig shared access to a single

computer was born lo companies to save costs [2]

In the 1970s, an operating system called the VM appeared, This operating system can run

on any virlual machine, has ils memory and other infrastructure, allowing data sharing

This makes the concept of virtualization in computers popular

In the 1990s, iclocom researchers and operators began to provide virtual private network

equipment, with good service quality and low cost This is a way for telecommunications

companies to provide multiple users with access to a single infrastructure

Another factor that also influences cloud computing is grid computing Grid computing allows major issues to be solved through parallel computing Computer utilities facilitate

computing resources to be provided as measurement and SaaS services that allow

subscribers lo be posted on the Internet and applications Since then, cloud computing

was officially bar Cloud computing was born af all these factors

We can see that within a minute, applications like Facebook have more than 4 million

new "likes", 300 hours of YouTube videos are posted, and more than 100 thousand calls

are made on the plalform Skype So what did they do to be able to store and serve such a

huge amount of data? ‘he obvious answer is Cloud Computing

With the development of cloud computing, ithas gained a lot of trust from organizations

of all types, sizes, and industries that are using this service for many use cases, such as

data backup, disaster recavery, cmail, virtual desktops, software development, and

testing, big data analytics, and interactive web applications with customers [3]

For cxample

A simple illustration to unravel the definition of cloud computing Before the 2000s,

users who wanted Lo use whal services had to buy the copyrighl, CD to install on personal

computers such as accounting software, address management, phone numbers, and staff

management This trend at that time proved to be quite cffective when just a few taps,

users can find out the necessary information about a certain employee or an address and phone number of an individual However, all dala is stored on a personal computer so (he mobility of these services is not high unless the user owns a laptop Nobody does that

anymore With just a PDA or an internet-connected personal computer, individual users

can search for essential information stored somewhere on the intemet There are many services about managing personal information, email on the internet that users can exploit such as Yahoo mail, Gmail, Facebook, Companies no longer have tn buy sofiware

licenses for employee management, financial management, no longer have to invest in

powerful compulers lo host the company’s dala Instead, the company only needs to pay

to use the service, more specifically, all the software for employee management and

financial management, installed by a service provider at a power center certam math

Servers with large amounts of memory are also installed by the service provider and the

company can stare the company's data Or as Health Care Companies are using cloud

k

services 1o develop more lailored (ealments for patients Financial services companies

are using cloud services to enhance detection and prevent scams in real-time

Besides that video, game makers are using cloud services to provide online games to

millions of players around the world

Thus, in the pasl, to deploy an application (for example, a Web silo), you had to buy/rent

one or more servers, and then set the server at data centers }, now cloud computing

allows you to streamline the buying renting process You just need to state your

requirements; the system will automatically gather free resources to meet your

requirements,

12 Components and structure of cloud computing

Cloud computing models include the Front end and Back end These two components are

connected via a nctwork, in most casos The front end is the transport vehicle through

which the user interacls with the system, the back end is lhe cloud The back delivers

applications computers servers and data storage to create a cloud of services ‘The front

consists of a business computer or network and applications used to access the cloud the infrastructure of today's cloud computing is a combination of reliable services delivered

by Lhe world’s leading information lochnology developers such as Microsofl, IBM,

Google based on virtualization technology (virtualized)[4] Cloud computing 1s divided into 4 basic layers with mutual intoraction including

¢ Cloud Application

The application layer of the cloud is to distribute software as a service via the Internet

Users do net need to install such applications on their devices Applications are easy to modify and users can easily receive suppor! fram the service provider

Key features of the application layer include:

- Activities arc managed in Lhe cloud (Internet), not on the cliont-side, Customers access

applications remotely via the Internet environment, the main interface is the website

- All operations such as version updales, patches are done in the cloud (service provider)

users do not need to perform these operations

© Cloud Programing

Tt provides a platform for the computing environment and solutions for computing

services, governs the infrastructure of cloud computing, and is the fulcrum for the

application layer to allow applications lo operatic on that platform Reduce the cost of

deploying applications These include Web 2.0, Mashups, and Concurrent and

Distributed

© Cloud infrastructure

Providing infrastructure of computers and devices in the cloud environment

(virtualization) Instead of customers having to invest money to buy servers, software,

connceted devices or rent physical infrastructure al data storage centers With the cloud

application, all customer requirements will be provided immediately with a much lower,

oven free investment This is also the most prominent feature in the cloud-computing

model

co Cloud Resources

Including the entire server, the connected device is designed and built specifically to

provide and operate cloud-compuling services Ofen-specialized devices with super-

powerful multithreading computing capabilities to meet the entire mining needs of a large

number of uscrs and their growing needs.

13 The Services of Cloud Computing

The components of cloud computing, such as front-end, back-end and cloud-based

distribution and network use Together, they form a complete architecture of cloud

computing In addition to the main components, such as SAAS, PAAS and IAAS, there

are all 11 components in cloud computing

However, in this report, I will only analyze three main types: IaaS, Paas and Saas,

13.1 SaaS: Software as a Service

Software is also known as cloud application service, this name represents many popular

enterprise services in the cloud market using the internet to contribute applications

Third-party management for its users Saa$ uses the inlernel lo deliver applicalions,

managed by third parties, to its users Most Saa8 applications run directly through the

web browser and do not require downloading or installing anything from the user

co The Distribution of SaaS

Because Saas's distribution model is web-based, SaaS eliminates the need for IT staff to

download and install applications on cach individual computer With SaaS, vendors will

manage all potential technical issues, such as data, middleware, servers, and storage,

allowing businesses to manage and support online

© The Benefits of SaaS

SaaS provides numerous benclits for cmployces and the company by significantly cuiing

down the time and money for boring tasks such as installing managing and upgrading

sollware sccurily This gives {cchnical sla{f more lime to focus on the company's urgent

issues

c The Characteristics of SaaS

‘There 1s a way to help you determine when businesses should use SaaS

» Scrvice management from a central location

+ Set on the remote server

* Accessibilily via the mlemet

+ Users are not responsible for hardware and software updates

co When to usc SaaS?

There are many separate specific situations that SaaS may be the best choice, including:

«Tf you are a startup or small company, you want to launch e-commerce services quickly

withoul need time and staff for server and software issues

+ For short-term projects that require cooperation.

* TÍ you use applications that do not require frequent use, such as tax software

+ For applications thal noed accessibility from the web and mobile

© The Example of SaaS

Google Apps, Cisco Webox, Drop box

13.2 PaaS: Platform as a Service

Cloud-based services, or Platform as a Service (PaaS), provide cloud components to

cerlain soflware while slill being uscd for core applications Users can use and build

custom applications thanks to the framework provided by the PaaS The enterprise or a third-party provider can manage all servers, slorage, and networking, while developers can maintain application management

Provides computing platform and a sct of multi-layer solutions Il supports application

deployment regardless of the cost or complexity of equipping and managing the

underlying hardware and software layers, providing all the features needed to support the

cycle Run fully by building and delivering an Internet-ready application and web service

without any download or installation of software for developers, TT managers, or ond-

users ILis also known by another name as cloud ware

‘The provision of platform services (PaaS) includes the conditions for application design, development, testing, deployment, and hosting of valuable applicalions such as

appheation services such as collaboration and packaging, web service integration and

integration, database integration, sceurily, scalabilily, slalc managemeal, applicalion

version, aud community benefits co-development and applied research ‘These services

are prepared as a web-hased integration solution

c The distribution of PaaS

PaaS's distribution madcl is similar to SaaS, except that instead of distributing software

over the internet, PaaS provides a platform for creating applications The platform is

~

distributed over the web and provides developers wilh free deployment, able to focus on building software white not having to worry about or worry about operating systems,

updates, or updates, Software updates, or infrastructure

PaaS onabled businesses to allow us to design and create applications that arc integrated

into SaaS with special software components, these applications or middleware are highly

scalable and available when they have special features Corlain cloud points

© The Benefit of PaaS

Xo matter what size your company is, there are many different benefits to using PaaS + Making developing and deploying applications easier and more cost-effective

+ Easy scalability

+ High availability

* Gives developers the ability to create custom applications without the hassle of

maintaining software

+ Significant reduction in the number of lines of code

* Automated business policy

* Allows easy integration with hybrid models

c The Characteristics of PaaS

PaaS has many features that define it as a cloud service, including

* Built on virtualization technology, meaning resources can casily scale up ar dawn as

your business changes

* Provides many services lo develop, test, and deploy applications

* Multiple usors can access the same service application al the same time

+ Integrated web services and databases

co When to use PaaS?

‘There are many cases when using PaaS is advantageous or even necessary if multiple developers are working on the same development project or together with other vendors, PaaS can provide great speed and flexibility to the whole process

You can completely create your custom applications, this cloud service can significantly

reduce costs, and it can simplify the challenges that arise if you require rapid

development or application deployment Thanks to the benefits of PaaS

© The Example of PaaS

AWS Blastic Beanstalk, Google App Engine, Windows Azure

1.3.3 Taa§: Infrastructure as a Service

Cloud infrastructure services, known as Infrastructure as a Service (Iaa8), are created

from highly sealable and aulomated computing resources Iaa8 is completely self-serving

to access and monitor everything like computers, networking, storage, and other services,

and it allows businesses to buy resources on demand and needs instead of having to buy the cntirc hard part themsclves

Stacking virtual archutectures is an example of a trend where everything is served and has somolhing in common Rather than a server for rent, contral slorage space or network equipment, workstations instead of investing to buy the resources, it is possible to hire

fall external services These services arc usually charged based on functionality and the

amount of resources used (and hence the cost) will reflect the level of activity Full of

development of web hosting solutions and virtual personal servers

The original name used was hardware service ([Taa8) and was created by a Nicholas Car economist in March 2006, but this is necessary However, the word was gradually

replaced by the concept of infrastructure services around the end of 2006

© The Distribution of laaS

TaaS distributes cloud-computing infrastructure, including everything like servers,

networks, operating systems, and storage, through virtualization technology These

specific cloud services arc provided lo the organization lhrough a control pancl or an

API, an IaaS customer has complete control over the entire infrastructure

TaaS provides technologies and capabilities as a traditional storage center, without having

to maintain or manage the hardware of all thosc components TaaS customers may still he

able to access their servers and storage directly, but external partners via virtual data

centers in the cloud hire all

Unlike SaaS and PaaS laa5 customers are responsible for application management,

runtime, operating syslem, middleware and dala Ilowever, IaaS providers manage

servers, hard drives, networks, virtualization, and storage Some vendors provide even

more serviecs outside of the virlualizauon layer, such as dalabascs

co The Benefits of TaaS

There are a large number of benefits to choosing IaaS such as:

+ Its the most flexible cloud model

+ Easily cnable storage deployment, networking scrvcr, and aulomated processing power

+ Additional hardware may be purchased based on usage

+ Provide customers with complete control of their infrastructure

* Resources may be purchased as needed

+ Rasy to expand high

@ The Charactenslics of IaaS

Few fealures lo consider when considering Iaa8:

+ Resources available as a service

* Costs vary depending on consumption

* Highly scalable services

* There are usually multiple users on a specific hardware

* Provide complete control of the organization's infrastructure

+ Flexible

© When to use IaaS

As wilh SaaS and PaaS, some specific situations are the best lime Lo use Iaa8 TaaS will

be the best choice because you do nat need to spend money and time trying to create

hardware and soflware if you arc slarlup or small company Iai is also beneficial for

large organizations that want complete control over their applications and infrastructure,

but they just want to invest in what is needed IaaS may be a good choice hecause you do

not need to upgrade or replace hardware or software to meet development needs for fast-

growing company Tt also helps if you are not sure about the requirements that a now

application will require because of ils flexibility in scalability or reduction as needed

c¢ The Example of JaaS

Digital Ocean, Google Compute Engine (GCI)

Conclusion: SaaS, PaaS and laas

In general, each cloud service model provides specific features and functions, and most

importantly, your organization needs to understand the differences between these

services Whether you're looking for a cloud-based software for storage purposes, a

platform that allows you to create custom applications, or want complete control of your

infrastructure wilhoul having to maintain your Physical component, there will be a cloud

service for you It does not matter which service you choose, but moving to the cloud is

the future for business and technology models, and it is necessary

1.4 Models of Cloud Computing

Cloud Computing has four service models (product models):

@ Public Cloud: services on the Cloud Computing platform to rent to individuals and

organizations, they share resources

a Private Cloud: used in a business and not shared with users outside the enterprise

0 Ilybrid Cloud: A hybrid model (hybrid) between the Public Cloud and Private Cloud

models

o Community Cloud: is a service based on Cloud computing platform built by companies

that work together and provide services to the community

Definition: a third party (seller) provides the services They exist cutside the company's

firewall and are managed by the cloud provider It is built for public use; users will

register with the provider and pay a fee based on the provider's pricing policy Public

cloud is the most commonly used deployment model of cloud computing

Target users: Including users outside the internet ‘The object of management is the

service provider

Advantages

= Serve more users, not limited by space and time

« The saving server system, electricity, and labor for businesses

Disadvantages

+ Suppliers-dependent businesses do nol have full management rights

+ Having difficulty storing internal documents and information

Ilowever, the Public Cloud has a problem, the data loss, and data security issue All data

on the Cloud service, which is protected and managed by the Cloud service provider in

this model This makes customers, especially large companics; feel inscoure about their

important data when using Cloud services

142 Private Cloud (“Businesses” Cloud)

Definition: Private cloud is the cloud computing services provided in businesses It is the

inevitable trend when these "clouds" exist within the company's firewall and are directly managed by businesses

Target users: Internal use and management by the enterprise

Advantages: Actively using, upgrading, managing, reducing costs, good security,

Disadvantages

© Difficulties in technology deployment and cost of system maintenance

c Restricting internal use within the enterprise, external users cannot use

143 Hybrid Cloud

Definition: A combination of private cloud and public cloud It allows us to exploit the

strengths of each model as well as offer optimal usage methods for users These "clouds"

are often created by businesses and their management will be split between businesses

and public cloud providers

‘Target users: Enterprise and supplier managed by agreement Users can use the service

provider and Jrom thers starl (heir business serviec

Advantages: Enterprises can use many services at the same time without limitation

Disadvantages: Difficult to deploy and manage It costs a lot of money

144 Communily Cloud

Community cloud (are services based on Cloud computing platform built by companies

that work together and provide services to the community)

LS The advantages/disadvantages of Cloud Computing and introduce some

applications of its

15.1 The advantages and Disadvantages:

© Advantages

‘Thus, in the past, to deplay an application (for example, a website), you had to buy/rent

one or more servers, then rent a server at the data centers, now cloud computing allows

you to reduce the buying / renting process You just need to state your requirements, the syslem will automatically gather free resources lo meet your requirements Therefore,

there are some basic benefits af cloud computing as follows

+ Use dynamic computing resources: Resources arc allocaled lo businesses exactly whal

the business wants immediately by mobilizing free resources available on the intemet

+ Reduce costs: Cosl reduction: Businesses will be able to cut costs to purchase, install

and maintain resources

Instead, they only need to identify their noods and thon ask the cloud service providers

to proceed

‘The amount of savings provided by cloud computing is significant Many businesses

report significantly reducing costs when switching to cloud computing Moreover, cloud

computing also saves time Many difficult tasks exist in the business world and with

different types of sofware, difficull tasks can be solved somewhal

+ Reduce the complexity in the structure of the business: Outsourcing the work of setting

up and operaling the IT apparatus, the business will only focus on the production of ils

specialized goods and reduce the complexity in the structure

| Increase the use af computing resources, slorage and expansion: One of the headache

questions of investing resources (ex: servers) is how long it will ran out of depreciation, Ï

invest as is it profitable or not, is it out of date on technology or not When using

resources on cloud computing, you no longer need to care about this anymore

Through the provision of soflware companics, customers arc provided free of charge wilh

cloud data storage An extremely large storage space This eliminates the need to

purchase related storage areas Moreaver, as this storage exists in cloud computing,

sorting and organizing become a simpler task

+ Mobility: Cloud computing exists within the limitations of the Internct As such, access

becomes much easier Your programs, files, and pholos slored in the cloud can be

accessed anywhere with an Internet connection In addition, these documents can be

accessed through methods other than a compulcr, such as a lablel or a smartphone

Through this mobility, emergency files you can gain access to any needed emergency

+ Durability: Cloud computing is also exlremely reliable, Usually, most saflware

development companies provide excellent customer support When the customer support

staff often has a high handling experience, troubleshooting is much less hassle

© Disadvantages:

While there are many advantages, however, cloud computing technology also has some disadvantages such as:

| Monthly cost: Probably the most prominent and common disadvantage for cloud

computing will be the monthly cost Using cloud computing usually requires bundled

sollware this can cost some extra software With cach custom program ercatod, the

cost will increase further

Some examples of ways to calculate costs in the Cloud

Fixed pricing: the supplier will specify the specification of a fixed computing capability

{allocated mcmory capacity, CPU type, and specd, ete.)

Unit pricing: popularly applied to the amount of data transferred, the amount of memory

allocated and used, ete ‘This method is more flexible than the above method.

Subscriplion pricing: most applications in the software service model (SaaS) users wall

predict cloud application usage narms (this is often difficult to achieve high accuracy)

| Dead time: All the time that needs to be taken care of is when the server downtime

This can occur for a variety of reasons such as weather updates, malware, or system

malfunctions If the backup sorver is nol working, your job could be severely hampered

Providers for cloud computing services will be difficult because of taking care of a large

number of customers at a time, With a large number, companies or servers can be

sometimes overwhelmed and some technical problems may occur

+ Compatibility: When integrating a completely new system, compatibility issues may or

may nol This is especially true thal companies want lo relain a program before exiting

Although it may appear fine and specific functions may result in a program failure,

resulling in compatibility or incompalibility issues occurs

+ Difficult to manage and secure: Because these systems often focus on the advantage of

sharing, sharing resources on the cloud, or the internet, From there, lo implement

management, security, and prevention of hackers is quite difficult Hackers can access

illogal information and stoal data, causing a grcat loss for the company Marcover, those

who have access to the server may use the information for malicious purposes In the

siluauon (hat the tacues of allack, sabotage of hackers lo profil more and more popular

In the future, we are sure that we will focus more on security for this cloud platform

+ Handling the data type of obsolete data: Data storage in the cloud system for too long is

one of the difficult problems Such as at Google driver Some user accounts store lots of files there However, these files have not been accessed or edited and used for many

years Combining so many types of data, our slorage space will be wasted Since then, many hosting companies often warn customers about the types of data that are outdated,

which they will delole in the future to avoid wasling slorage Tesour

Conclusion: Cloud computing offers a multitude of advantages and many

disadvantages In most cases, however, companies and businesses will find

cloud computing to be extremely useful for their development and maintenance

with this advanced technology.

15.2 Some applications of Cloud Computing

- Cloud database

- Testing and development

- Saving

- Big Data Analysis

- Slore and share dala

- Backup and restore data

- Enterprise management application

1.6 Typical cloud applications

More and mors companies participating in the development of typical cloud applications such as Microsoft, Google, Intel, and IBM have been creating a large market of cloud applications The research team would like lo offer a few Lypical cloud applications:

16.1 Google App Engine

Google App Engine (G.AE) allows us to deploy our application on Google's

infrastructure Building applications with App Engine is casy, convenient during

maintenance, easy to expand when there is increased traffic, or additional storage needs

GAK supports us with 2 application development environments: the Java runtime

environment and the Python runtime environment

Java execution runtime environment: You can build applications based on Java

technology including the Java virlual machine (TVM), the Java servlet, and the Java

programming language - or any language compiled on the JVM such as JavaScript or

Ruby

Python Runtime Environment: is a specialized environment including a very fast

interproter and a standard library

Some features of Google App Engine:

+ Support the most popular, most popular web technologies

* Automatic expansion and load balancing

* Provide APT to authenticate users and send mail using Google accounts

+ Provide a development environment with full features

1.6.3 Amazon Web Services

Amazon Web Services is a collection af services that provide programmers with access

to Amazon's ready-to-use computing architecture infrastructure Well-built computers

that have been built and refined over the years by Amazon can give anyone acccss ln the

Intemet Amazon provides several Web services that meet some of the core requirements

of most systems: storage, computing, messaging and data sets

Amazon provides standard interfaces based on SOAP and REST to interact with each

service Libraries developed from Amazon or from one of the permissible languages,

such as Ruby, Python, Java, Erlangen and PIP, to make exchanges wilh these services Command-line touls can also perform computational resource management on EC2 ‘The

REST inicrlacc is relatively casy to use; You can usc the client-side program writlen in

any language as the HTTP protocol to allow making requests to Web services,

Storage

Amazon Simple Storage Service (S53) provides Web service interfaces for data storage

and recavery Data is given in any kind and can be stored and accessed from any location

via the Internet You can store an unlimited number of objects in S3 with the size of each object ranging from 1 byte to 5 GB The data warehouse can be located in the United

States or European Union countries You can choose the storage location for your objects

when you create buckets (similar to the concept of a directory) Data is securely stored using the same Amazon storage infrastructure

According to the leading experts on cloud computing, the development of cloud

computing in the future will facus on three mam issues including: Likability (Federated),

automation (Automated) and Client aware (terminal aware) These are also new

approaches to TT automation that allow uscrs to respond to user requests in a new, more

efficient and cost-effective manner Linked clouds will allow faster arrangement of

resources, while endpomt-aware clouds will take advantage of the unique capabilities of

18

cach device in an optimal way Cloud computing will be the most used technology in the

future

1.6.2 Windows Azure

Windows Azure cnabics the deployment of Windows applications and data storage on

Microsoft infrastructure via the Intemet environment

Windows Azure provides an application development cnvironment using NET

Framework, Native Code Support for common languages like C #, Visual Basic, C | 1,

ar maybe by ava Use Visual Studio or another development tool

Both Windows Azure applications and on-premise applications can access Windows

Azure storage services through the REST API The storage service allows storing large data objects via blob, provides queues for communication belween components in the application, and provides Table-based data with a simple query language (SQL) For

applications thal require traditional data storage, the Windows Azure Platform provides

SQL Azure database services

Every application has a configuration [ile in Windows Azure, By configuring this Lile, the

appleation manager can configure the components that Windows Azure applications

need to run

Computer Services

Windows Azure computing service can run many different types of applications The

main goal of this archilecture is o support applications that have large amounts of

concurrent traffic Windows Azure 1s designed to support running muluple copies of the same application on different virlual servers

1.7 The situation of apply Cloud Computing in the world and Vietnam

Criginating from grid computing from the 80s, on-demand computing (Utility

computing} and service software (SaaS), Oracle is a pioncer in deploying this technology

So far, cloud computing is being developed and provided by many vendors, including

Amazon, Google, Dala Synapsc, Sales force as well as traditional vendors Microsoll,

IBM, IP Has been a lot of people Personal use to big companies like L’Oreal,

General Electric, KRay, Coca-Cola accepted and used

Research firm Gartner assesses that the chief priority of CIOs will be virtualization and cloud computing enterprise applications to help their companies less worry about

managing (heir inirastruclure believe in; focus more on slcoring the company's

development process According to the assessment, as of 2012, 80% of enterprises in the

list of the top 1,000 companies (according to Fortune magazine - USA) will use at least

some types of cloud services and about 20% the business will no longer own IT assets or

infrastructure

Intending lo solve dala problems, calculations, services, for customers, cloud computing

has been bringing great profits, providing, a playground and a large market for suppliers scrvice, so ils rapid growth can be counted Zom day lo day

According to VNUnet newspaper (April 28, 2009), 27.7% were concerned about data

slorage scrviecs in a survey of customers using cloud computing services; 17% on

financial related applications, 12.8% on e-mail

A recent report by the Gartner rescarch team highlights the development of Cloud

Computing in the context of globalization According to the survey, about 50% of

businesses and organizations in the world are applying (his modem Lcchnology wilh a

growth rate of 17% per year In 2011, global cloud computing revenue reached US § 2.4

billion Gartner predicts this number will nearly quadruple in 2012

Not out of the trend, Vietnam is gradually approaching cloud services through projects of some foreign enterprises such as Microsoft, Intel as well as from domestic developers and suppliers such as FPT, Biaki IBM is the pioneer enterprise in launching cloud

computing center in Vietnam in September 2008 with the first customer is Vietnam

Technology and Communication Joint Stock Company (VNTT) It can be said that

Vietnam is one of the first countries in ASEAN to use cloud computing,

‘This technology is considered a solution to the problems that many companies are facing

such as lack of IT capacity; limited infrastructure investment costs Currently many

companies are wasting resources such as not fully exploiting the public server system

capacity, investing loo much in human lenms Meanwhile, in theory, cloud compuling

will allow businesses to not focus too much on infrastructure or application upgrades, do

not require large human resources and can easily change scale as needed

Although cloud computing is currently widely used in many developed countries around

the world because of the significant benefits (hat il brings, bul in Vietnam, businesses are

not interested in this technology However, according to experts, this is the optimal

solution for businesses in our country to reduce costs as well as increase productivity at

maximum level

Regarding the current situation of cloud computing application in Victnamese

enlerprises, the following conclusions can be drawn: Currently, there are a few large

enterprises in Vietnam that put cloud computing into application and business

performance significantly improved However the number is quile small Most still just

stop at interested and inquiring

So what arc the main reasons for barricrs to access and Iransition to the environment,

such delayed cloud computing technology:

The first and most important reason is the perception of the business Fach enterprise has

a different scale, organizational model, operating in different fields, but the guidelines and policies must be smoothed from the management tcam, cxcculive board to the IT

team and Relevant departments in the business Because, when selecting and switching to

this technology platform, they face huge challenges in terms of capital, budget for initial

investment costs (if building the Private Cloud or Ilybrid Cloud model }, costs of

modernizing production and business processes, business management and operation

processes, elc technical barriers, information securily Ii can change all the traditional processes that businesses operate

The

ond reason is transmission bandwidth Cloud computing is highly dependent on

the speed and bandwidth of the connection, directly affecting the performance and

response time of the application for the end user This is also one af the bottlenecks

affecting the quality of service

‘The third reason is the applicalion, data and information security issues when everylhing

of an enterprise is put to the cloud ‘This is also a matter of great interest and discussion in

recent years: legal corridors, mechanisms, policies and methods to prevent and protect these important applications and data

However, it is also undeniable thal the role and value of Cloud Computing is a major

factor m promoting business transformation process in businesses and at the same time

helping businesses operate more efficiently and intelligently More and more cost

savings

Currently the situation is improving significantly According to a recent survey of

Symanlec, a leading software company in Lhe world, about 46% of Vietnamese

enterprises and organizations are deploying cloud computing, and other virtualization

projcets

Security firm Symantec also said that Vietnamese businesses have a special interest in

cloud computing applications and the opportunilics that this now lechnology brings

Results from the survey show that 39% of domestic enterprises are currently using

private virtual software (VPS) services, while 21% are virtualizing servers and databases

The application of cloud computing technology in business as well as in life is an

incvilable development step with the trend of the times It is predicted that this is the 3rd

wave of technology, which will affect the habits and thinking of applying technology

today The difficult thing is how businesses as well as individuals accept the trend

For most Vietnamese businesses today (mostly small and medium-sized businesses), the

language barrier is the biggest obstacle in the search for technologies to assist them

Therefore, Vienamese suppliers will be a bridge for domestic enterprises lo follow the world technology trend It requires a certain level of technology, an understanding of the

habits and culture of Vieinamesc businesscs, and above all the ability to tram the market

associated with technology Cloud Technology Provider in Viemam does all three things

well, so the Vietnamese market will not only be a potential market anymore

Cloud computing technology is a common trend of the era, the application; widespread

development is indispensable, especially in the context of globalization But to keep up

2a

the trend, to market flourish, there are many urgent things Changing a working mindset,

an operating habit is somethmg that suppliers have to do Vietnamese enterprises see and

accept.

H Big Data - Overview And Its Impact Gn Health Care

IL1 The definition of Big Data and What is Big Data in IIealth Care System and

Why we need it?

ILL.1 The Definition Of Big Data

Big Data is the term has meaning a very large and complex data set that the traditional data tools cannol process [owever, Big Data contains a lol of valuable information that

if successfully extracted, will help a lot for business, scientific research, predicting

upcoming discascs, and even Real-time traflic conditions delrmination Therefore, these

data must be collected, organized, stored, searched, and shared in a different way than

normal In this article, invite you ta learn about Big Data, the methods penple use to

exploit it, and how it helps our lives

According to the page SearchNataManagement "Big Data is a combination of structured,

scmi structured and unstructured data collected by organizations thal can be mined for information and used in machine learning projects, precictive modeling, and other

advanced analytics application”

‘The tremendous pragress of science and technology, to the advent of communication

channels that require an cxtremely large global data systcm such as social nclworks and

advanced technology devices, poses a challenge It is not smali for different industries to

find other ways to handle data

Big Data (big data} is a collection of data that has a capacity beyond the performance of

traditional applications and tools Big Data size is getting bigger and mare important in

the modern world and especially the current industry 4.0, as of 2012 it can range from a

few dozen terabytes to many petabytes (1 petabyte — 1024 terabyte)

IL1.2 What Is Big Data In Health Care System? And Why We Necd Il?

The application of big data in health care has many impacls It will use specific data of patients at the time of examination and it can help to prevent disease, cure diseases, and

reduce costs, cic

Doclors always want to know as much as possible about a palient, and any signs of a

disease as soon as possible to treat any disease in the early stages it will always be

simpler and less expensive, as well as have a better cure for incurable diseases

With today's improved technology, collecting that huge data has become easier and it is

also converting il inlo impartanl, uscful, rolovant information, which can then be used to

provide for better care

This is the purpose of health-care data analysis: using data-based findings to predict and

solve problems before it's too late, but also to assess treatment methods and regimens

quickly better, better daily tacking, involving patients mare about their own health and

empowering them to access tools to de so

IL2 The Characteristics Of Big Data

Tn 2001, analyst Doug Laney of META Group (now the research firm Gartner) said

that the challenges and opportunities in data growth can be described in three

dimensions: increasing volume, increase in velocity (velocity) and increase in variety

(variety) They along with many other companies and organizations continue to use this "3V" model By 2012, Gartner added that Big Data in addition to the above three features must "need new forms of processing to help make decisions, delve mls

things, and optimize the working process" Here is Gartner's new concept of Big Data

2014 under the "5s" model - 5 important propertics that say Big Data

o Volume Talk about the amount of dala ercated and stored The size of the data will

be assessed as valuable and potential, and to consider whether it can be considered as

about here, it can easily process and store all the information on separate small branch

systems

o Variety (Variety of data types) This concept refers to the type of data (nature of

data) This helps analysts effectively use the details of the results They are composed

of text (lext), image (image) sound (sound), video, plus il completes the missing part

through data aggregation algorithms

9 Velocity (High-speed processing capability) In this day and age, data rates are

created and processed to meet the needs and challenges that lie in the path of growth

and development ig data is usually available in real-time [6]

© The speed of data created and then transferred from one place to another has now

reached a staggering rate Like people can chat with each other on Facebook at a fast pace in loday's network cnvironmenl, Big data allows us lo analyze the generated dala

without saving them to the database

o Variability: The current data types are se diverse and inconsislenl that il can

interfere with its processes and management

ø Valuc of data The quality of the data collected can vary greatly, which will greatly

affect the accurate analysis We can see this is nature as well as the concept that

businesses or rescarchers who want lo use and cxploil Big Data must hold and

understand it first

13 Stored in Big Data Analysis System

11.3.1 Hard disk storage technology

Disk storage often uses low-cost hard drives for long-term storage Disk storage can be

done via distributed file systems or dalabases

“ Distributed file storage system

Distributed file systems support schema-less data storage, providing redundancy and high

availability by copying data to different locations implemented with a distributed file

syslem thal provides fast access, capable of storing large data sets with semi-structured

and unstructured data and it provides fast read / write capabilities , solving the velocity

characteristics of Big Data

A distributed file system is not ideal for datasets with large numbers of small files

bocause this oroalcs excessive disk scarch, slowing down the overall dala access process

Due to these limitations, distributed file systems work best with fewer but larger files

processed sequentially Many smaller files are aften combined into a single large file to

allow optimal storage and processing

Distributed file system storage device is suitable for large data sets of raw data In

addition, this is an inexpensive storage oplion for storing large amounts of data [or

extended periods of time that need to be maintained online This is because it is possible

lo add hard disk sets lo the cluster without offloading data to store data oflinc It should

be noted that distributed file systems do not provide the ability to search the content of files as a slandard [7|

* RDBMS relational database storage system

RDBMS (Relational DB management sysloms) are very good al handling transactional

workloads involving 4 small amount of data with random read / write properties

RDBM8s adhore to the ACTD design principle, so RDBMS systems (limited by a node}

do not support redundancy and fault tolerance

‘To handle large volumes of incoming data at a fast pace, relational databases often need

to scale [lowever, RDBMS systems only have the mechanism of vertical division, not

scaling horizontally, which is a big drawback This makes the RDBMS not ideal for long-

torm data storage accumulated over time Relational databases, on the other hand, often require data to conform to the schema Therefore, RDBMS systems that store semi-

structurcd and unstructured data must perform indirectly This results in high latency,

which makes the relational database not ideal for high-speed data storage while requiring

a high-availability database storage device wilh fast dala writing capabililics

Due to these disadvantages, traditional RDBMS is often not as useful as the main storage

device in Big Data solution environment

+ Non-SQL database storage system

The emergence of Non-SQL storage systems mainly meets the characteristics of Big Data

(Volume, Velocity, and Varicty) The slorage requirement of increasing dala volume

requires the use of highly scalable databases while reducing the costs for businesses to

remain competitive Non-SQL slorage systems mect this requirement by providing

scalability while using inexpensive commodity servers

Fast data flow requires a database with fast access data logging capability Non-SQF

slorage systems allow [ast wriling using schema rules when reading inslead of writing rules on the schema Having high availability, these systems can ensure latency when

node / network errors occur The storage system needs lo handle different dala formals

inchading documents, emails, images and videos and incomplete data Non-8QI storage systems can store different lypos of semi-structured and unstructured data In addition,

this system supports non-schema data

Based on the lype of dala stored, Non-SQL dala storage systems arc divided into 4

categories: Key-Value, Document, Column family and Graph data storage systems

+ NewSQI database storage system

Non-SQL storage devices are highly scalable, available, fault tolerant and fast for read /

write operations However, they do not provide the same transaction and consistent

support as shown by ACID-compliant RDBMS Under the BASE model, Non-SQL

storage devices only provide consistency in the final processing state instead of

immedialc consistency Therefore, they arc nol suilable for use when implomenting large-

scale trading systems

NowQuery slorage devices combine RDBMS's ACID design principles with scalabilily

and loss tolerance

11.3.2 Slorage technology in memory (In-Memory Databases)

A device in memory often uses RAM, the main memory of the computer, as a storage

medium ta provide fast data access Increasing capacity and reduced RAM costs, along,

with the increasing read / write speed of hard drives, have helped develop solutions for storing data in memory Stormg data in memory eliminates 1 / O latency of the disk and data transfor time between main memory and hard drive This overall reduction im road / write data latency makes data processing much faster The capacity of in-memory storage

devices can he massively inorcased hy expanding horizontally of the cluster that is

storing the storage device in memory

Cluster-based memory allows storing large amounts of dala, including BigData dala sels,

which can be accessed significantly faster when compared to disk storage devices This

significantly reduces the overall execution time of Big Data analysis, thus enabling Big

Data analysis in real lime

A sequential read of 1 MB of data from a storage device in memory takes about 0.25 ms, while the same amount of dala from a disk slorage device takes about 20 ms This proves

that reading data from internal storage is about 80 times faster than storing it on disk

In-momory storage device allows in-memory analysis, uscd lo analyze data in memory

Tn-memory analysis allows for operational analysis and operational BI through rapid

execution af queries and algorithms

Basically, in-memory storage allows us to understand fast data flows in Big Data

environments (velocity characteristics) by providing storage modia that facilitate the

creation of detailed information over time real This supports quick decision making to minimize threats or take advantage of opportunities,

Mass Data Storage device is deployed on a cluster, providing high availability and

redundancy ‘Therefore, horizontal scalability can be achieved by adding more buttons or memory When compared lo on-disk storage devices, in-memory storage devices are

expensive because of the higher memory costs compared to disk-based storage devices

Although a 64-bit machine can use 16 exabyles of memory, due Lo ils physical

limitations, such as the number of memory bays and memory installed, it 1s significantly

less For scaling up, it is not just the addition of additional memory but also the addition

of nodes that are required eleven memory limits per node This increases the cost of data

slorage

In addition to being expensive, m-memory storage devices do not provide the same level

of support for durable data storage The price factor further affects the achievability of

the device in memory when compared to the disk storage device Therefore, only the

most updated and the most valuable data is stored in memory that old data is replaced

with newer, newer dala

Depending on how it is implemented, an m-memory storage device can support

schomaless storage or schema awareness slorage Non-schema support is provided

through key-value data storage

Storage devices in memory can be deployed as: IMDG slorage lechnolegy and IMDB

storage technology Although both of these technologies use memory as the basis for data

storage, whal makes them different is the way they store data in memory

In-Memory Data Grids (MDG) technology stores data in memory as key value pairs

on multiple nodes where keys and values can be any business object or application dala Which application is in sequential form? This supports storing non-schema data through semi-structured or unstructured data storage

+ IMDB (In-Memary Datahases) technology uscs database technology and takes

advantage of RAM performance to overcome the runtime latency issues that cause

disk storage devices

IMDB is able to store structured data (relational IMDB) or can take advantage of Non-

SQI technology (non-relational IMDB} to store semi-structured and unstructured data

Unlike IMDG, which provides access to data via API, IMDB relational using SQL

language should be more [amiliar, supporting data analysts or data scicntists without

advanced programming skills high Kon-SQL-based IMDBs often provide API-based