A Comprehensive Solution to Evaluate the Performance of Openstack

Cloud Platform for E-Governance

Dr. Bhushan Jadhav

, Dr. Arun Kulkarni

, Dr. Nikhilesh Joshi

and Prof. Sonali Jadhav

Department of Artificial Intelligence and Data Science, Thadomal Shahani Engineering College, Bandra, Mumbai, India

Department of Information Technology, Thadomal Shahani Engineering College, Bandra, Mumbai, India

Department of Computer Engineering, Thadomal Shahani Engineering College, Bandra, Mumbai, India

Keywords: Cloud Computing, E-Governance, Big Data Analytics, Openstack, Hadoop, DevOPs.

Abstract: The evolution in Information and Communication Technology (ICT) has made enormous digitalization in

every one of the parts over the globe. In E- governance, this digitization has profoundly changed the method

for connections among citizens and governments. Because the government is responsible for delivering basic

contributions and administrations to its citizens and other stakeholders, modernization is essential for

improving the administration and efficiency of government activities. The traditional E-governance lacks in

computational speed, storage strategies, automation and intelligence in deployment of applications. As a

result, modern technologies such as cloud computing, big data analytics, and DevOPs can help them satisfy

the exponential demand for e-governance services while also providing additional benefits that were never

conveyed. The Openstack is a popular cloud computing platform that has built-in Big data analytics support

and DevOPs automation solutions that can meet current e-governance requirements while overcoming the

existing challenges. As a result, the purpose of this research paper is to provide a methodology for leveraging

the benefits of cloud computing with integrated big data analytics and DevOPs by implementing them over

the Openstack cloud platform. This paper has covered all the important aspects of deploying Openstack for

E-governance that leverage built-in Big data analytics and DevOPs to facilitates the colossal benefits which

were never before. Finally, the paper also covers the implementation and performance analysis of Openstack

cloud platform for E-governance using MaaS and juju along with the benchmarking results.

INTRODUCTION

The E-governance is nothing but digitalized

provisioning of government services over the web

that has been expanded powerfully to the public in

getting access to their services and products [1]. The

main focus of E-governance should be efficient

service delivery of E-governance portals and

applications to each and every stakeholder stays in

urban and rural areas for making E-governance

popular among stakeholders. The benefits of E-

governance lead to an increase in transparency,

effectiveness, less corruption and the convenience of

access at low cost [3]. The present E-governance isn't

always being broad next to startling mark because of

limited funds, infrastructure and other insufficiencies

which made the impact on its usage [4]. So, in the

beginning, the usage of E-governance was quite

limited, but today it has increased, possibly due to

https://orcid.org/0009-0000-6456-0515

https://orcid.org/0009-0008-7699-8272

various schemes of governments and population [5].

Despite developments in digitization, present

technologies are unable to match the current demand

for E-governance. So the government needs to look

about technological improvements to satisfy the

existing demands at lower costs [2].

Cloud computing is a distributed computing

approach that allows users to access a shared pool of

resources, applications, and services from any

connected device, at any time [8]. Rackspace and

NASA developed Openstack, an open-source cloud

architecture, to promote cloud standards and lays a

strong foundation for cloud development. Cloud

computing is emerging and demanded technology to

address existing problems with computing, whereas

big data analytics can overcome constraints in data

acquisition, storage, processing, and analytics [6]. It

is the most extensively used technology for creating

private and public clouds; hence it has the greatest

Jadhav, B., Kulkarni, A., Joshi, N. and Jadhav, S.

A Comprehensive Solution to Evaluate the Performance of Openstack Cloud Platform for E-Governance.

DOI: 10.5220/0013344400004646

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 1st International Conference on Cognitive & Cloud Computing (IC3Com 2024), pages 115-122

ISBN: 978-989-758-739-9

115

community of developers and contributors who

utilize it. [18]. The Openstack and Hadoop are Open

source Cloud computing and Big data analytics

technologies, which are analogous to each other in

most of the aspects. As both the technologies can be

run on commodity hardware and compatible with

each other, they can be integrated together for saving

the infrastructure. As a result, same hardware can be

used for performing cloud computing as well as

processing Big data over the virtual clusters provided

by Openstack cloud [14]. The unmatched benefits

provided by integrated Openstack with Hadoop and

DevOPs would be mechanism for data capture,

storage and processing thus saves the capital

expenditure on infrastructure and man- power,

facilitates huge storage and highly scalable clusters

for data processing etc.

Therefore, the aim of this research paper is

to propose an Openstack-based model for E-

governance that includes all facets of a modern

computing environment. The re-search paper has

presented the way to implement and integrate

different E-governance services on the top of

Openstack platform along with complete validation

and testing. Finally, the benchmarking results of

Hadoop deployed on Openstack using E-governance

data and DevOPs tools deployed on Openstack for

automation have been presented. Upon

implementation of proposed model, it is expected that

the integrated approach of Openstack cloud platforms

for e-governance can provides many features such as

integrated DevOps tools, software- defined

datacenter operations, software-defined storage,

software-defined networks, self-healing for failures,

automated service provisioning, integrated

monitoring of services, high availability, disaster

recovery, load balancing, auto-scaling, high security,

efficient data processing with insightful analytics, and

graph-based analysis which were never before.

RELATED WORK

Many researchers have already researched and

proposed the need of using cloud computing in E-

governance projects. The overall research focuses on

justifying the need of technology improvement in the

domain of E-governance. Some of them are discussed

as follows.

Muzahidul Lalitha, Bhavani Jivandham et al. [6],

speaks about the adoption of ICT that can

dramatically reduce the corruption from a society and

improve the govern-ance. Tamara Almarabeh,

Yousef Kh. Majdalawi, & Sasikala P, have

emphasised the necessity of E-Governance and how

cloud computing can provide integration

management with automated problem resolution,

control end-to-end security, and assist in budgeting

based on actual data usage. On a global scale, cloud

designs can help governments eliminate duplicative

operations and boost resource efficiency [5, 7]. Pusp

Raj Joshi, Shareeful Islam et al., has proposed the

strategic framework for integrating cloud computing

in E-governance applications [10]. John Carlo Bertot,

Heeyoon Choi, emphasised the promises and

capabilities of Big Data in transforming digital

government services by governments, as well as the

importance of the interaction between governments,

citizens, and the business division in moving from

Smart government to Transformational government

[11]. Arun.J Mohamed and Hazaruthin.M suggested

the advantages of cloud computing and big data in

enterprise applications. They have also outlined two

unmistakable and highly clear patterns that are

defining enterprise computing, revealing a

tremendous lot of potential for a new era of integrated

applications [12]. According to Shubham Awasthi

and Anay Pathak, the Openstack cloud platform is

massively scalable and multi-tenant, with numerous

linked services that manage storage, compute, and

network resources in the data centre. These services

work together to provide an IaaS. The application

programme interface (API) facilitates the

integration of several services [15]. H. T.

Ciptaningtyas, R. R. Hariadi, and Shubham

Awasthi et al. have revealed the insights of the

Openstack cloud platform, which is massively

scalable and multi-tenant, with numerous

interconnected services that manage storage, compute,

and network resources in the data centre [15], [19].

Jadhav B., Patankar A., Shubham Awasthi et al.

discussed the several services launched with the

Juno version of Openstack, including Nova, Swift,

Horizon, Glance, Neutron, Cinder, Heat, Keystone,

Ceilometer, Sahara, and Trove [6],[15].

In light of the research presented above, no

researcher has explored the adaptability of cloud

solutions for E-governance, as well as their

evaluation for production deployment. As a result, the

purpose of this study is to propose an Openstack-

based model for E-governance that incorporates all

features of a modern computing environment, as well

as to examine its verification and validation utilizing

the Openstack cloud platform.

METHODOLOGY

The methodology for proposed research paper

carries out different methods to vali-date the

IC3Com 2024 - International Conference on Cognitive & Cloud Computing

116

successful implementation of proposed cloud

computing and big data analytics enabled E-

governance. The methodology involves the following

aspects.

3.1

Proposed Model

The proposed model of E-Governance is made up of

seven separate layers. Each Layer contributes

distinct capabilities to the development of cloud and

Big Data enabled E- governance. The proposed

model's architecture follows a bottom-up approach,

with the lower layer supplying functionality to the

upper layer, as illustrated in Figure 1. The proposed

Model has seven layers, which are discussed below.

The bottom layer of the proposed architecture is the

Technology Infrastructure Layer, which outlines the

use of infrastructural components such as servers,

networks, and storages to build high- performance

data centers that will be used for automated cloud

service creation and delivery. A data center is a

facility that houses IT hardware such as computing

devices, storage systems, and networking

equipment. Its primary purpose is to keep IT

resources up and running with little downtime. [9].

This layer also creates a software-defined data center

that leverages a bare metal hypervisor to automate

virtual resource deployments.

The Openstack Cloud Deployment Layer is in

control of getting resources from the technology

infrastructure layer and building an Openstack cloud

on top of that [6]. It acquires resources like CPU,

Memory, IO, Storage and network from virtualized

re-source pool provided by SDDC. The Openstack is

built over the supported hypervi-sors like KVM and

Xen.

The proposed model's Big data analytics layer is

in responsible for collecting, storing, processing,

managing, and analysing the massive amounts of data

created by various E-governance websites. The

proposed approach incorporates a big data analytics

layer into the Openstack cloud platform using the

Openstack Sahara Project. The Sahara project is the

Hadoop data processing module within the

OpenStack that enables organisations to create

Hadoop clusters or run Big Data analytics

applications in the cloud [7]. Various NOSQL

databases can also address the Big Data problem. The

Openstack Trove component includes many NOSQL

databases for capturing, storing, and processing

unstructured data supplied by various E- governance

websites [17].

The DevOps layer provides various tools to make

collaboration between development and operation

phases in software development life cycle. DevOps in

Openstack can fulfill the different stages of software

development such as planning, coding, building,

testing, releasing, deploying, operating, and

monitoring by providing tools such as GIT, Jenkins,

Puppet, Ansible, Saltstack, Chef, Docker, Nagios,

and Kubernetes [6]. In E-governance, these DevOps

steps can help construct sophisticated software

products such as web portals and applications fast and

easily. The public network layer makes the

communication between different stockholders of E-

governance using internet while User access layer

allows different stockholders of E- governance like

citizens, employee, enterprises and other Government

departments to access the E- governance services and

applications by means of using the different delivery

channels and user access devices like Desktop,

laptop, mobile, tablets, kiosk etc.

Figure 1: Openstack based proposed model for E-governance.

The Inter-Operability Framework layer is responsible

A Comprehensive Solution to Evaluate the Performance of Openstack Cloud Platform for E-Governance

117

for delivering different E- governance services to the

various stack holders by means of service delivery

gate- ways. It uses various Inter- operability standards

provided by middleware technologies, web services,

SOA, micro Services and REST.

The integrated approach of Hadoop and

Openstack cloud platform for E-governance gives

many features like integrated DevOPs tools, Software

defined datacenter operations, Software defined

Storage, Software defined network, self- healing for

failures, automated service provisioning, integrated

monitoring of services, high availability, disaster

recovery, load balancing, auto scaling, high security,

efficient data processing with insightful analytics and

graph based analysis etc.

3.2

Experimental Setup for

Implementing Proposed Model

The experiment for implementing proposed model is

conducted on a DELL Studio Workstation Server

XPS 9250 with Intel Core m5 X980 processor at 3.3

GHz, 256 GB RAM, Seagate 250 GB SSD, SATA

3Gb/s disk, and 1 Gbps network connection [6]. Here,

Ubuntu Linux version 21.10 64-bit distribution is

used for the deployment of Openstack Mitaka release

with the help of MaaS and juju. [9]. Metal as a Service

(MaaS) is used for data centre management (DCM)

by automating MaaS in combination with Juju, can

easily model and deploy complex environments and

has optimized provisioning for production hardware.

The default services deployment of Openstack MaaS

using juju is shown in Fig. 2. Once Openstack is

deployed the required VMs for Web server,

Application Server, Database Server etc. can be

deployed using Openstack dashboard Horizon. The

Hadoop clusters can be created by installing Sahara

Component over the compute node using Juju. On the

top of Hadoop using HDFS, the E-governance multi

variety data can be restored for testing purpose. In this

way, the proposed model can be deployed over

Openstack.

RESULTS AND DISCUSSION

The performance of proposed model over the

Openstack is tested over the three test cases. The test

case one calculates the response time and average

speed over the standalone machine and Openstack

VM, test case two calculates the execution time

between Hadoop deployed on Standalone system and

Openstack Sahara cluster with same configuration

while test case three test the performance between

two databases deployed on Openstack like Mysql and

Hadoop Hbase by comparing the execution time and

average latency between them.

4.1

Test Case 1

The compute performance of Openstack is tested by

calculating response time and average speed over the

physical server and VM over the Openstack server

[13]. For testing the performance the hardware

configuration used for both are kept same. The

response time of the proposed system is calculated

by recording the time taken by standalone machine

and Openstack VM to perform read operation with

different file size shown in Fig. 3.

Figure 2: Deployment of Openstack MaaS using juju.

From Fig. 3, it is seen that the Virtual machine

running over Openstack gives better response time in

performing read operation over different file sizes.

The average speed is calculated by counting the

IOPS by transferring the file be- tween standalone

system and Openstack based cloud Virtual machine

which is shown in Fig. 4.

From Fig. 4, it is seen that the average speed for

performing IO operations over Openstack cloud is

performing well in all the aspects.

IC3Com 2024 - International Conference on Cognitive & Cloud Computing

118

Figure 3: Response time for dedicated hardware and VM

Openstack.

Figure 4: Average Speed for dedicated hardware and VM

Openstack.

4.2

Test Case 2

In this test case, the performance of Hadoop system

over the proposed model is tested by calculating the

execution time for two Map- reduce programs like

Word count and Sorting. The execution time is

calculated by recording the time between Hadoop

running on dedicated hardware and Hadoop running

on Openstack cloud platform with same hardware

configuration [14].

The execution time for the sort program is

recorded with three input datasets of size 100 Kb, 1

Mb and 100 Mb. The time command is used to

calculate the elapsed time of a map- reduce

operations. The output of above command calculated

over the dedicated Hadoop deployment is shown in

Table 1.

Table 1: Execution time for sorting by dedicated Hadoop

deployment.

Similarly, the execution time for Hadoop over the

Openstack cloud in proposed E- governance model

with same hardware configuration is recorded which

is shown in Table 2.

Table 2: Execution time for sorting by proposed Hadoop

deployment.

The Fig. 5 Shows the comparative analysis of

execution time between dedicated Hadoop

deployment and Openstack integrated Hadoop

deployment in proposed model.

Figure 5: Evaluation of execution time for sorting.

From Fig. 5, it is concluded that for smaller input

dataset the dedicated deployment works well but as

soon as the data size grows the dedicated Hadoop

takes more time than the Openstack integrated

Hadoop. Likewise the execution time is calculated for

Word count program with input size 100kb, 200kb

and 5000kb.

The output of Execution time on dedicated

Hadoop deployment and Openstack inte- grated

Hadoop deployment in proposed model is shown in

Fig 6.

Figure 6: Execution time for Word count app by Dedicated

and Openstack Hardware

4.3

Test Case 3

In test case 3, the rural health statistics dataset is

A Comprehensive Solution to Evaluate the Performance of Openstack Cloud Platform for E-Governance

119

utilized to compare the performance of MySQL with

Hadoop's NoSQL database, Hbase [17]. The data set

includes healthcare records of numerous residents

from rural India [10], [17].

The first test measures the execution time in

seconds for MySQL and Hbase by querying a

healthcare dataset with varying numbers of records.

In MySQL, the show profile command is used to

record the execution time. In Hbase, the execution

time is presented by default in the query results [17].

Fig. 7 shows a comparison of execution time in Hbase

and MySQL . Fig. 7 shows that the MySQL database

takes significantly more execution time than the

Hbase database.

Figure 7: Execution time in Mysql and Hbase.

The second test records the average latency

generated by Mysql and Hbase databases using

Mysql's built-in performance schema and Hbase

latency command given below [17].

$hbase .regionserver.fsReadLatency_avg_time

Fig. 8. Evaluation of Average Latency in Mysql and Hbase.

Figure 8 shows the comparison of average latency

between Hbase and MySQL. It shows that the

MySQL database has a higher average latency than

Hbase [17].

4.4

Benchmarking Results

The benchmarking results of Openstack cloud

platform are calculated over bench- marking tool

Rally installed on the controller node of the

Openstack. The Rally is an Openstacks Open source

benchmarking tool that automates and monitors the

Open- stack deployment for benchmarking &

profiling [11].

Figure 9: Rally deployment command on Openstack.

Figure 10: Response time for boot and delete instances.

It tells you how Openstack per- forms in a load at

scale. It can be used for validating the performance

tests and benchmarking the deployment over the

pluggable benchmark scenarios provided by Rally

[12]. The Rally has big ecosystem of cooperative

services which notifies when something fails or

performs slowly or does not scale [13]. The Rally

benchmarks can be calculated over the Cloud

compute node using various built-in scenarios. To

de- ploy a benchmark, the rally deployment create

command is used which is shown in Fig 9.

The Rally task start command calculates the

response time for deployed scenario with parameters

like boot server and delete server instance is shown

in Fig. 10. After executing the task start command,

the benchmarking report in html format is generated

and can be seen using task report command shown

in Fig. 11.

IC3Com 2024 - International Conference on Cognitive & Cloud Computing

120

Figure 11: Rally Benchmarks for Nova Instances.

Similarly detailed statistics for creating and booting

a nova instance can also be seen using rally start over

soon as shown in Fig 12. The benchmark for above

scenario is depicted in Fig. 12.

Figure 12:

Rally statistics for creating and booting a

nova instance.

Figure 13: Rally deployment scenarios.

Likewise for each scenario the benchmark results

can be calculated. The different benchmark

scenarios provided by rally are depicted in Fig. 13.

From above benchmarking results, it has been

observed that, Openstack cloud platform gives better

benchmarking results than the other private/public

cloud platforms in terms of load testing,

responsiveness, storage deployments and the

scalability.

CONCLUSION

In this way, it is concluded that the Openstack based

proposed model for E- governance benefits

governments in all the aspects. The coupling

between Openstack cloud platform and Hadoop

framework gives the unique solution for E-

governance practices and offer plethora of features

which can be customized profitably. The efficiency

and effectiveness of the proposed model is tested by

implementing it over the MaaS and juju deployment

of Openstack. The deployment results using

different test cases conclude that response time and

average speed of Openstack cloud platform for E-

governance gives faster computing power as well as

Hadoop deployment gives minimal latency and

better execution time than existing deployments. The

benchmarking results demonstrates the Openstack

for E-governance is faster in provisioning and

releasing cloud resources using faster boot and

delete response time. By implementing the Cloud

enabled E-governance, Government may get

tremendous accomplishment in every aspect on

account of unexceptional advantages like cost

reduction in service delivery, improved interaction

between the government and the stakeholders,

eliminated manual processing in government offices

by automation, faster service delivery, efficient

processing of applications, location independence,

reduction in corruptions, and ease of access. These

benefits of E-governance attracted most of the

governments in developed and developing nations to

leverage it for effective government operations.

Finally, we can conclude that the Openstack cloud

platforms for E-governance can provides many

unique features which were never before such as high

availability, disaster recovery, load balancing, auto-

scaling, high security, integrated DevOps tools, self-

healing for failures, efficient data processing with

insightful analytics, and graph-based analysis etc.

A Comprehensive Solution to Evaluate the Performance of Openstack Cloud Platform for E-Governance

121

REFERENCES

1. Brahimi, M.: Towards a based-agents and web services

framework for managing Algerian e- government

applications. Int. J. Electron. Gov. 14, 437-457 (2022).

2. Dash, S., Pani, S.K.: E-Governance paradigm using

cloud infrastructure: Benefits and challenges. In:

Computational Modeling and Security 2016, pp. 843-

855. Elsevier (2016).

3. Suri, P.K., Sushil: Strategic Planning and

Implementation of E-Governance. Springer,

Heidelberg (2017).

4. Mohammed, F., Ibrahim, O., Nilashi, M., Alzurqa, E.:

Cloud computing adoption model for e-government

implementation. Sage Journals (2016).

5. Balkrishna, A., Singh, S.K., Pathak, R., et al.: E-

governance paradigm in the Indian agricultural sector.J.

Discov. Agric. (2024).

6. Jadhav, B., Patankar, A.: A framework for integrating

cloud computing and big data analytics into e-

governance using Openstack Sahara. In: Information

Communication Technologies for Intelligent Systems

2018, Smart Innovations, Systems and

Technologies,pp. 1-13. Springer, Heidelberg (2018).

7. Chen, C.L.P., Zhang, C.Y.: Data-intensive applications,

challenges, techniques and technologies: A survey on

big data. Int. J. Inf. Sci. (2014).

8. Almarabeh, T., Majdalawi, Y.K.: Cloud computing of

e-government. Commun. Netw. J. 8 (2016).

9. Jadhav, B., Patankar, A.: A novel solution for cloud-

enabled e-governance using Openstack: Opportunities

and challenges. In: Computers Network and

Communications 2018, Communications in Computer

and Information Science. Springer, Heidelberg (2018).

10. Joshi, P.R., Islam, S., Islam, S.: A framework for cloud-

based e-government, from the perspective of

developing countries. J. Future Internet (2017).

11. Bertot, J.C., Choi, H.: Big data and e-government:

Issues, policies, and recommendations. In: Digital

Government Research 2013, pp. 1-13 (2013).

12. Mohamed, A.J., Hazaruthin, M.: Analytics as a service

delivery model for the cloud. In: Engineering and

Technology (ICETECH) 2015, pp. 1-5. IEEE (2015).

13. Tornyai, R., Pflanzner, T., et al.: Performance analysis

of an OpenStack private cloud. In: Cloud Computing

and Services Science 2016, pp. 282-289 (2016).

14. Corradi, A., Foschini, L., et al.: Elastic provisioning of

virtual Hadoop clusters in OpenStack-based clouds. In:

Communication Workshop (ICCW) 2015,pp. 1914-

1920 (2015).

15. Awasthi, S., Pathak, A., Kapoor, L.: OpenStack -

Paradigm Shift to Open Source Cloud Computing & Its

Integration. In: 2016 2nd International Conference on

Contemporary Computing and Informatics (IC3I),pp.

112–119. IEEE, Greater Noida, India (2016).

https://doi.org/10.1109/IC3I.2016.7917944.

16. Ciptaningtyas, H.T., Hariadi, R.R., Husni, M., Ghozali,

K., Sholikah, R.W., Setyadharma, I.M.D.: OpenStack

implementation using multinode deployment method

for private cloud computing infrastructure. In:

Intelligent Technology and Its Applications (ISITIA)

2023, pp. 12-17. IEEE (2023).

17. B. Jadhav, A. B. Patankar, S. B. Jadhav: A Practical

approach for integrating Big data Analytics into E-

governance using Hadoop. In. Second International

Conference on Inventive Communication and

Computational Technologies (ICICCT) 2018, pp. 1952-

1958.IEEE(2018).

18. Binod Kumar Adhikari, Wanli Zuo, Ramesh Maharjan:

A Performance Analysis of OpenStack Cloud vs Real

System on Hadoop Clusters. In: 2017 IEEE

International Conference on Edge Computing (EDGE),

pp. 194–201. IEEE, Honolulu, HI, USA (2017).

https://doi.org/10.1109/IEEE.EDGE.2017.34.

19. Sasikala, P.: Cloud Computing and E-Governance:

Advances, Opportunities and Challenges. Int. J.

Comput. Appl. Comput. 2(4), 32–52 (2012).

https://doi.org/10.4018/ijcac.2012100103.

IC3Com 2024 - International Conference on Cognitive & Cloud Computing

122