Infrastructure of Cloud Networks

ABSTRACT

Networking Technology is evolving at fast pace from simple internet connections to present day ubiquitous computing. Today, networks have developed to an extent where concepts of parallel computing, grid computing and the most recent cloud computing is possible and popularly used. Cloud computing is an emerging field in networking technology which was developed to improve the overall IT infrastructure. Cloud networks focus on tackling a variety of key issues of computing. Any cloud network involves multiple computers that are connected through a real-time communication network. Cloud computing shares number of resources such that coherence and economies of scale is achieved.

This paper covers the infrastructure of cloud networks, various ways in which it can be deployed and used for different purposes, the security aspect of cloud networks and its future scope along with the challenges it might face.

INTRODUCTION

Cloud networking is very much related and goes hand in hand with cloud computing. The basic concept of cloud includes, providing on-demand access to resources via network.

Cloud networks were developed to provide different organizations or businesses and users access to centralized computational resources. In the recent years, the introduction of cloud has changed the way one values companies. Cloud penetration has exploded, and software and virtualization have taken precedence over physical networks and on-premise deployments of hardware.

While shifting to cloud networks, organizations are going step by step & assessing applications as they go. They also have to take into consideration the performance, risk factors, integration and costs during the process.

In a cloud enabled network, the core infrastructure like routing, packet forwarding etc. remains in the premises but the issues such as network management & monitoring, the security services, data, maintenance etc. are done through cloud. Cloud-based networking manages and monitors the resources. It provides connectivity between the resources and the applications deployed on cloud as and when required.

Cloud Network Architecture[1][2]

 

Figure I

Every cloud networking architecture consists of the following components:

1)       Front End

2)       Back End

Both ends are connected to each other through internet via routers.

Front End

The front end is that part which includes users and devices which is client side.

It consists of interfaces along with applications which are quite important to access cloud computing sub-system.

Back End

 

This section consists of cloud sub-system

It consists of all components and resources which are important to provide cloud-based services to users.

Data storage, Virtual Machines (VMs), Virtual Storage, Security, Cloud based services, deployment models, servers etc.

Security, Traffic Control and protocols happens in back end.

Components inside the cloud infrastructure[3]

 

1) Hypervisor

       It is a low-level program which is basically a firmware that acts as a Virtual Machine Manager (VMM).

       The main task is to share physical instances of cloud resources between several customers.

2) Management Software

       The main function is to maintain and configure the infrastructure.

3) Deployment Software

       Assisting in terms of deployment and integration of applications on cloud is done by using (DS).

4) Network

       Delivery of cloud services can be done both publicly and privately by using internet.

 

Layers and Services of Cloud Computing[4]

 

SaaS i.e. software as a service provides cloud web applications to the users. Here, the need to download an application and then use it is eliminated. The users can directly go on to their web browsers to use this service. SaaS providers handle or take care of all the technical issues such as data, storage, servers etc. The most commonly used example of SaaS is Gmail. The server here is hosted by Google and the users are just accessing the service or their e-mails through the browser as a client.

Few more examples are: Dropbox, Salesforce, Cisco WebEx, Google Apps etc.

Figure A1

 

Instead of directly providing software or an infrastructure to control everything, like the name suggests, PaaS (platform as a service) offers a platform to develop your own application. This service is offered over the web and a user can develop his/her application online. Here, the user doesn’t have to concern himself with the OS used, the software updates, storage etc. The user must create his/ her application and then just deploy it on cloud using some specific steps which vary from service provider to service provider. The runtime, middleware, OS, virtualization, servers, storage, networking aspects of the application are taken care of by PaaS. The most important advantage of using PaaS is: it allows users to develop applications without worrying about the maintenance of it. 

A few examples of PaaS providers are: Heroku, Google App Engine, OpenShift, AWS Elastic Beanstalk and Windows Azure.

 

Figure A2

 

 

 

The main motto of IaaS (information as a service) is to provide maximum control to the clients using this service. It is mostly designed for applications or businesses where the clients don’t want to invest in buying hardware for services such as storage, networking etc. IaaS allows users to configure their and or outsource their own virtual data centers. This will allow different businesses to scale up or down the resources as and when required. In this type of service, the application, data, runtime, middleware and OS level configurations can be done by the users. This is possible using the APIs offered by the cloud service providers. IaaS is the most flexible model compared to the rest of the services. It is a self-service which allows its users to monitor and buy resources (such as computer, networking storage etc.) on-demand. Since resources are available as a service and the cost depends on how much a user is using, spending extra on unnecessary hardware requirements is prevented.

The examples of IaaS providers are: Google Compute Engine, Microsoft Azure, Amazon Web Services, Rackspace, and Digital Ocean etc.

Basic structure of IaaS consists of:

       Hardware

       Data Centers

       Bandwidth

       Load balancers

       Virtual Server Space

       Cloud Hosting

       Firewalls

 

                                Figure A3 

 

 

Data as a service (DaaS) is a cloud computing model which provides and distributes information in the form of data such as text, images, videos, audios etc. which is made available to customers opting for service via internet. This data is provided to user irrespective to geographic separation of receiver and sender. With development and improvement of IT infrastructure and the increasing need for data lead to development of DaaS concept. The actual platform on which data resides is of less importance due to Software Oriented Architecture (SOA). Because of increasing demand for technologies like IOT, Big Data, NFC, sensors etc., more data is being generated which needs to be stored and provided as service. Various data sources can be mobile devices or smart phones, sensors embedded in different products or physical things, social media content (such as tweets, texts, posts, blogs etc.), data from internet search history, customer transactions and the transactions that occur within different banks as well.

 

Figure A4: DaaS Architecture

 

Deployment of Cloud Computing[5]

The cloud infrastructure is developed in such a way that it can be exclusively used by a single organization comprising multiple consumers (e.g., business units). It can even be owned, managed, and operated by an organization, a third party, or some combination of them, and it may exist either on or off premises. 

The cloud infrastructure can also be used for open use by the public. It can be owned, managed, and operated by a business, academic, or government organization, or some combination of them.

The cloud infrastructure can be explained as a comparison of two or more distinct cloud infrastructures (private, community, or public) that are unique entities, but are bound together by standardized or proprietary technology. The hybrid model allows the data and application to remain portable (one example would be: cloud bursting for load balancing between clouds).

 

The cloud infrastructure is also designed such that it can be common to a specific community of customers or users from different organizations that have shared concerns or motives (a few examples would be: mission, security requirements, policy, and compliance considerations). A community can by itself manage and operate by involving one or more of the organizations in the community, a third party, or some combination of them, and it may exist on or off premises.

 

 

Cloud Computing Security[6]

 

Cloud networks have a lot of advantages which has made huge organizations move their data into cloud. This shift to cloud networks has raised the security concerns. Organizational data, user data etc. could be anywhere on cloud, and once it becomes a hotspot for hackers, all the credentials, important details are in trouble. 

Cloud networks are prone to various attacks such as:

 

Data breaches: These are very common, since the data is spread out across regions, finding hotspots to get important data is easy. Data breach happens when malicious users get a hold of actual user information.

Denial of Service Attacks: When services are made unavailable to the users, its known as Denial of Service. Organizations are supposed to provide a few services, and when these are made unavailable, it may result in the customers losing trust.

Compromised credentials: Data breaches lead to this threat. The main reasons why data breaches may happen in the first place are poor passwords, bad key or certificate management etc. One way to prevent this would be Multifactor authentication (asking for one-time passwords…)

Malware injection: Malware injection is when a malicious piece of code injects itself within the actual running software. This is a threat because; it may provide hackers unauthorized access to data due to this code.

Vulnerabilities in APIs & browsers: Users or clients of cloud use cloud services through browser APIs. It acts as a medium between the services and users. If these APIs or browsers are compromised, cloud networks are too. How secure an API to the cloud is defines the security of a cloud network.

Shared Vulnerabilities: Keeping the cloud networks safe and secure lies in the hands of the clients too. And hence, provider and consumer cooperation are a must to maintain the security of cloud.

 

Key Industrial Organizations[7]

 

According to an article published by Forbes in November 2017, the organizations that are achieving milestones in cloud networking are:

Microsoft- Microsoft has its involvement in all three service aspects of cloud (IaaS, PaaS & SaaS). Windows Azure is the cloud service provided by Microsoft. It allows the clients to deploy ML/ AI applications in innovative production environments.

Amazon- Amazon’s AWS (Amazon Web Services) is the cloud service provided which is impressive and also available online for a free trial so that users can test whether or not it is suitable for their applications.

IBM- IBM is one of the top 3 companies because it has helped organizations shift their software’s from on-premises to cloud-based applications. IBM also has its services in all three deployments of cloud (IaaS, PaaS & SaaS)

Salesforce- Salesforce is enormously successful in the SaaS business and one of the reasons why it is in the top 10.

SAP- SAPs HANA technology is the reason why it ranks in the leading businesses in cloud. HANA has been deployed across thousands of organizations.

These were the top 5 leading businesses in this sector. A few more organizations that are doing well rank as follows:

Oracle

Google

Service Now

Workday

VMware

 

 

Use Cases of Cloud Networking

 

Extend the existing applications: One of the most important advantages of cloud is its scalability. Cloud allows the application to be scaled up or down as and when required.

 

Cloud hosted desktops and applications: Cloud networks allow the applications to be developed and deployed on cloud. Their maintenance, computing etc. is then managed by the cloud provider.

 

Disaster Recovery: This is one of the most important IT aspects for any software product. Cloud network’s flexibility, capacity & cost makes disaster recovery one of the use cases of it.

 

Future Scope and Challenges[8]

 

There is no one possible future for cloud. Each sector will benefit from it using different possible solutions and hosting their own hybrid environments. But the one that is in the picture right now is pushing data and communications to the edge of the cloud. Since the most sophisticated applications of Internet of Things require computing on edge this will be the nearest future of cloud networks. With 5G in picture, pushing these communications to cloud and their computation seems more than possible.

But along with this, there will be a raised concern for major issues such as security, latency etc. Keeping the networks safe and secure will be one of the most important challenges faced.

CONCLUSION

 

The conclusion that can be drawn after all the research is that, cloud networks have revolutionized the way networking and different organizations work. The “on-demand access” and “pay for what you use” policies have made computations, storage of data, and maintenance & monitoring of networks much cheaper, elastic and easier. Cloud networking has brought a lot of advantages along with it and made developing applications & deploying them on to cloud for maintenance and monitoring easier as well.

But along with this, the security concerns have increased to a higher rate. Not just that but since computing takes place on cloud, the latency and communication speeds are also a problem. With edge networking next in the picture, these problems and concerns will only increase. Since, computing is so much cheaper now; edge computing is a possibility but will face greater security concerns. Low or almost negligible latency though seems possible due to 5G, achieving it in real time is still a difficult task.

REFERENCES

TY – BOOK, AU – Jadeja, Yashpalsinh,AU  – Modi, Kirit,PY  – 2012/03/01,T1  – Cloud computing – concepts, architecture and challenges, DO  -,10.1109/ICCEET.2012.6203873,ER  – Cloud computing – concepts, architecture and challenges

https://www.researchgate.net/publication/254035330_Cloud_computing_-_concepts_architecture_and_challenge

Cloud Computing Architecture

http://communication.howstuworks.com/cloud-computingl.htm

Eugene Gorelik, “Cloud computing Models”,

Working Paper CISL# 2013-01, January 2013

Available: http://web.mit.edu/smadnick/www/wp/2013-01.pdf

Stephen Watts, “SaaS vs PaaS vs IaaS: What’s the Difference and How to Choose”

Available:

https://www.bmc.com/blogs/saas-vs-paas-vs-iaas-whats-the-difference-and-how-to-choose

TY – BOOK, AU – Lee, Young-Chan,PY  – 2012/08/09,SP  – T1  – A Deployment Model for Cloud Computing using the Analytic Hierarchy Process and BCOR Analysis

Available: https://www.researchgate.net/publication/259972968_A_Deployment_Model_for_Cloud_Computing_using_the_Analytic_Hierarchy_Process_and_BCOR_Analysis

V. D. Sharma, S. Agarwai, S. S. Moin and M. A. Qadeer, “Security in cloud computing,” 2017 7th International Conference on Communication Systems and Network Technologies (CSNT), Nagpur, 2017, pp. 234-239.doi: 10.1109/CSNT.2017.8418544keywords: {cloud computing;security of data;computing environment;cloud computing paradigm;economic benefits;security issues;security concerns;Cloud computing;Security;Computational modeling;Companies;Virtual machining;Computing Paradigms;Cloud Deployment Models;Privacy;Hypervisor;Encryption},URL: http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=8418544&isnumber=8418499

Bob Evans, “The Top 5 Cloud-Computing Vendors: #1 Microsoft, #2 Amazon, #3 IBM, #4 Salesforce, #5 SAP

Available: https://www.forbes.com/sites/bobevans1/2017/11/07/the-top-5-cloud-computing-vendors-1-microsoft-2-amazon-3-ibm-4-salesforce-5-sap/#4ebc7f996f2eMcheick, Hamid & Obaid, F & Safa, Haidar. (2012). Cloud Computing: Past, Current and Future. Available:

https://www.researchgate.net/publication/267250165_Cloud_Computing_Past_Current_and_Future