Evaluation of Cloud Networking Design, Security, and Architecture Infrastructure - Coursework Example

?Cloud Networking Evaluation of Cloud Networking Design, Security and Architecture Infrastructure. Miss Ranna Almoussa 1128267 Computer Science and Technology University of Bedfordshire Luton, UK, 2013-14 Email: ranna.almoussa@study.beds.ac.uk Abstract: The adoption of cloud computing concept into IT networking is nothing but ‘Cloud Networking’. This service builds up cooperative transmission, communication and cloud data accessing through cloud networking. Cloud Networking reduces overall operational risk, corporate capital expenditure, networking complexity and maintenance cost. This new concept improves networking functionality, flexibility and enhances dynamic flexibility to virtual resources; it also increases the Virtualization and optimization for example. It decreases latency or network load. However, this new concept has many challenges’ in future. This research paper presents the new concept of “Cloud Networking” and its design architecture, sustainability and security in the networking infrastructure. The primary goal of this report is to explain the challenges, solutions and need that are associated with cloud networking. Key words: Cloud Networking, sustainability, Design, Architecture and Infrastructure. I. INTRODUCTION Cloud computing is a very broad term and includes a variety of aspects and dimensions, which are very helpful in hosting a large number of applications and making them available to any device at any time [1]. Mostly it is used by business houses whether they are big, small or medium size. But the concept first originated back in the days when cloud computing was introduced by companies like Google, Yahoo, Amazon, etc. and now is used on a large scale in organizations like Dell, IBM, HP, Microsoft, and alike. The emergence of cloud networking has eliminated the concept of ‘silo computing’ where an application or software was of restricted use and available to only selected systems. A cloud networking system creates multiple layers and multi-tiered systems inside a cloud server, allowing for enhancement of cloud protection and security and application of complex software [2]. The report aims at understanding the development of cloud networking, its impact on the environment and how various organizations joined forces to formulate plans to protect the natural surroundings. II. DEVELOPMENT OF CLOUD NETWORKING There has been a paradigm shift in how computers are used to store, manage and retrieve data in an organization. The very concept of cloud computing has satisfied the uneven demands of users who wanted cheap computing resources which was able to record, store, retrieve and disseminate information. Cloud networking poses some new challenges when compared to conventional OSI systems. Many new components and dimensions have been added to the use of computers, the internet is one of them. The cloud network now includes a layer of web applications, which is additional in comparison to traditional OSI systems [15]. Cloud networking works with IaaS (Infrastructure as a service) and SaaS (Software as a Service) while aiming at reducing consumption of energy and cost, leading to a green cloud. Though in early times, sleep mode was seen as a very effective way of decreasing energy usage, but now as servers are required to be online 24*7, expenditures made for maintaining them have increased by many folds; cloud networking is the solution for it. The concept behind the development of cloud networking is using Software as a Service (SaaS). There are a large number of servers in a data centre and several processors intensive software which requires more processing power, in which case they need more developed networks which are capable of handling big data storage capacity and transferring capability [20]. Figure 1: Software as a Service (SaaS) [15] A. ATTRIBUTES OF CLOUD NETWORKING Scalability: There are a large number of nodes that are involved in a cloud network. It is necessary that they must measure the rate of successful message and information delivery to the user so that it can be ensured that they do not become a bottleneck. The network fabric must be such that it is able to handle the load of throughput of facts to numerous packets [2]. Low Latency: Cloud networking has decreased the latency time, i.e. the time that users had to wait for the information be delivered to them. Superior networking makes sure that data is made available to the user almost instantly. The cloud networks have almost microsecond latency resulting in improved application performance and utilization of the server. Self-Healing and Resistance: Since the servers and data centers have to work on a constant basis, there is no room for downtime or shutting down for maintenance [5]. Cloud networking has self-healing capability and software updates. These networks have the ability to update themselves on a regular basis, identify the problem and rectify it almost immediately. Extensible Management: Managing and ensuring upgrade of the system is a very exhausting and complicated process. There is a risk that the system may collapse. Cloud networking helps in upgrading the system across all networks [7]. Figure 2: Cloud Computing[17] III. Data Centers The proliferation of internet and e-commerce content has been the mainstay and feature of the computer era. They have influenced the storing of data in a very big way thereby working of data centers. The infrastructure required of these centers also changed dramatically, with more emphasis now being given to the development of networks that are capable of healing themselves and take very less time to deliver accurate information to users. There are three main types of data centers (DCs): private, public or virtually private. Private DCs are to be used by one entity only, as they are owned by it [10]. They are connected to internet via secure gateways and passwords and use internet in a very limited way. On the other hand Public DCs are connected to the internet and are used for web-based services. It includes multi-tenant services and commercial public services as well. While Virtual private DCs build on common data center infrastructure which are often complimented by the cloud service provider in the form of IaaS. IV. Data Centers and Cloud Networking Virtualization of storing and processing facilities has enabled the DCs to provide an environment that offers resources and applications to multiple tenants at the same time and at the same infrastructure. These tenants may be different organizations/enterprises [11]. Computing Virtualization has made it easy for tenants to access the data that they need to. Through such a process virtual machines (VM) can be created on the server with a dedicated Central Processing Unit (CPU) cycles, memory, storage and bandwidth. In order to support a large number of such tenants and facilitate networking among them, intra and inter-DC networks can be used which have evolved from the basic Virtual Local Area Network (VLAN). A tenant network service can be virtual Ethernet LAN (Local Area Network) service, an IP VPN or an internet service. The data centers that may be connected to one or more network service providers to ensure connectivity to users accessing the DC [21]. The intra-cloud and the inter DC network can be obtained as a leased fiber or private line service. It can also be an Ethernet VPN service or an IP VPN service. V. Network as a Service: Quality of Services (QoS) – AWARE CLOUD NETWORKING SERVICES The network as a Service (NaaS) has become increasingly important and relevant, as selection of network of cloud providers is completely dependent on the needs of the application. The lack of QoS on cloud networking has significantly impacted the application performance. NaaS proposed a cloud model which presented a new and fresh on-demand cloud networking service by changing users as per the ‘pay-as-you-use’ model [13]. It deals with connectivity management. For instance, Amazon propose the Virtual Private Cloud (VPC) which provides a secure network to enterprises related with Amazon. Cloud Networking provides a framework for building a network which is on-demand and is given importance according to their type (gaming, instant messaging), importance (bank transactions, emergency) , etc. Recent studies fortify the fact that current cloud computing applications suffer from poor QoS. It significantly impacts performance of the application. There are two factors that supplement the QoS adoption of NaaS model: absence of suitable networking model services; and lack of differentiation in the class of services (CoS) to cloud applications [25]. The Service oriented Approach (SoA) is now seen as a very important part in cloud networking. It presents an analytical model to evaluate QoS in the network. Cloud networking has gained all the more important because firms like Microsoft recently have proposed to build private virtual networks in their cloud so that end-to-end sites can be connected together and the flow of information can be made simple and smooth. Cloud networking not only ensures that the QoS requirements of cloud applications are met, but also the selection of adequate network services [24]. While cloud computing states that various services must be assembled together, it poses another challenge in designing of network services. One of the main issues presents itself when functional and non-functional aspects of QoS of cloud networking are separated from one another. In NaaS, the network is required to regularly upgrade itself as the hosted cloud application can change its requirement over time. If the network is not kept updated, QoS will degrade due to network over-provisioning, disruption, etc. It also does not take into account requirements of the different applications flow. For instance, real time applications require low latency, whereas data intensive applications need high throughput. To combat these issues, a QoS model can be suggested here which will link properties of QoS in each of the virtual nodes [23]. This model follows four criteria: Availability (A): It is the portion of time that is available to Cloud Networking Service (CNS) to complete the requested service without any failure. Reliability (R): It is the rate of conformity as per the offered service in comparison to demanded service. Capacity (C): It is the competence of CNS to fulfill the required service. It can be calculated using different parameters such as Bandwidth, transaction number, etc. [12]. Delay (D): It can be said as the total time taken by CNS to complete its task. It can be characterized by many different attributes like latency, processing time, etc. VI. SECURITY ARCHITECTURE FOR CLOUD NETWORKING Main reason which describes as to why cloud networking has been developed can be evaluated from the fact that it provides a whole new level of security to cloud computing. Security and protection is one of the main aspects to cloud networking [17]. It offers reduced capital expenditure, operational risks, maintenance and increased scalability. Cloud networking adds another dimension to cloud computing in the form of networking. It enables dynamic and flexible placement of virtual resources and optimizes performance of the cloud in the form of reduced latency, or network load. In recent times, as cloud computing has gone viral; the infrastructure for hosting the clouds has changed completely. Necessity of networking has increased many times as the servers are required to stay online and operational all the time, 7 days a week, there is no option as to whether they can be stopped or shut down for maintenance [22]. The emergence of cloud computing has given birth to a new requirement, i.e. security infrastructure. Such networking has made protection and safe keeping of the data and information in clouds. One of the major feature of cloud computing is that it involves sharing and transfer of knowledge and facts, as a lot of virtual resources are moved from one system to another. Therefore it is necessary that proper security checks are kept on the activities of the cloud, in order to assure that operator’s infrastructure follows the service user’s demands. The service provider can change the placement of virtual resources, measure the security demands according to security functionality, and if necessary move it to another infrastructure provider [26]. Safekeeping is one of the major challenges in cloud computing, and the service user has to ensure that those requirements on cloud infrastructure are fulfilled. There are many security parameters that are needed to be fulfilled, and networking makes sure that they are met. Only after a positive evaluation, does a service user moves his data or processes on to the virtual infrastructure. In case if these are not fulfilled, service users can remove his information from the provider’s infrastructure. There are various security issues that the cloud service providers face; like ensuring that their infrastructure is up-to-date and see that their client’s data is kept safe [1]. There are many cloud security architectures like Deterrent Controls: which are put in place to prevent any purposeful attack on the cloud system; preventative controls: which upgrade strength of the system by managing the vulnerabilities; Corrective controls: these are used to reduce the impact of a hacking or a security breach attack; detective controls: to detect any malicious activity in the infrastructure [3]. There are numerous dimensions to cloud security which states that correct security controls should be kept in place but in accordance to an asset, threat and vulnerability risk assessment matrices. Some of the dimensions are identifying management; physical and personnel security; privacy; business continuity and data recovery; logs and audit trails; public records, etc. Cloud computing offers many benefits, but at the same time it is very vulnerable to threats [4]. To minimize and decrease the risk, cloud stakeholders must invest heavily in improving the cloud networking. It will help in assessing the risk and ensuring that systems are properly encrypted to protect the data; building higher assurance into auditing to strengthen regulatory compliance. VII. THE SUSTAINABILITY AND SURVIAVBILITY NETWORK DESIGN There are a range of issues that are associated with the use of cloud computing; like its negative effects on the environment; though it is often considered to be a green form of computing [6]. The primary sustainability issue that is concerned with cloud computing is the energy use. Clouds are being used in the most extensive way as most of the world population is now using cloud to share their information and knowledge across various platforms. There are various ways in which it is being used like in gaming, email, social media, mobile applications, etc.; therefore it can be said that impact of cloud computing on the natural environment is very widespread [14]. Though cloud service providers have made significant investments in data center infrastructure so that not only raw computing facility and power but also SaaS business applications can be provided to users. But they have not been at all helpful and effective in reducing the carbon emissions or its carbon footprint [19]. Numerous models have been developed and designed that work to ensure sustainability and survival of cloud computing and networking in the long term, like Energy-Aware Dedicated Backup Protection (EABP) which provides two paths for traffic flow: primary and link-disjoint backup path. In it the back-up path remains online until traffic flow is either completely terminated or reduced to such a level which the primary path can handle without nay hassles. In this model there is no Sleep mode for the system. Another model is Energy-Aware Backup Protection with the sleep mode. This mode here acts as hibernation mode, where the infrastructure’s most resources get into a resting mode when there is no traffic; but can be re-activated in a short-time when the need arises [16]. It is used only for the backup path and resources, but only if they are not blocked by active traffic. Though these two models consider energy consumption and do their bit to cut down on it, but are not very effective. A third model can be introduced here ‘Energy-Aware Shared Backup Protection (EASBP)’ with the sleep mode. It is very effective in finding a trade-off between energy efficiency and backup capacity minimization [18]. There are many instances, where cloud service providers have collaborated with one another to work towards building such approaches and techniques to reduce negative impact of cloud computing on natural surroundings. For example, Greenpeace and Facebook declared that Facebook would start to use clean and renewable sources of energy to carry out its operations and day to day activities [9]. Similarly Apple Inc announced that it will make all of its data centers ‘coal free’ and increase the use of solar energy. Also Microsoft and Accenture decided to work together to assess the environmental impact of IT operations like Cloud Computing and Networking. VIII. REGIONAL FAILURE AND SURVIAVIBILITY FOR CLOUD NETWORKING SERVICESS Cloud networking consists of various computers that reside in various geographical areas but connect with each other’s and allows users to store, manage, process and retrieve data. All the computers are connected to each other with a real time communication network. There are various fundamental models have been developed for facilitating the cloud networking [8]. These service models define the type of services that is provided by the cloud network to its users. The platform as a service which is also known as PasS is a type of service model which is used in the cloud networking. Under this model the user develops software by utilizing the tools and library which are supplied by the provider. The tools supplied mainly include the networks, servers and storage [23]. The main feature that PaaS offered to its user is that users can develop software with the use of various applications without incurring any costs of purchasing and managing the hardware. PaaS service has various vendors who provide unstable scalability and maintenance level. Apart from the above mentioned service the PaaS also consist of various other facilities such as designing of computer application, development and testing of computer application. It also consists of team collaboration, integration of databases, security and storage of databases, versioning of applications and integration of web services. PaaS have been growing rapidly and has been adopted by business organization for making their activities and processes quicker and faster [2]. PaaS offering also consists of a mechanism for managing various services that include the management of workflow, monitoring and reservation. Use of cloud networking has changed the application development process significantly. This concept of cloud networking has delivered new ways of creating business applications. The PaaS facilitates the company for running application without bothering the maintenance of hardware and infrastructure of software. The PaaS work along with the infrastructure as a service and software as a service. Under the PaaS service model the cloud networking provides the user with a platform which can be used by the developers for building applications and service over the internet [10]. Companies that had been working on the older technologies have now adopted the platform as a service for building various business applications. Cloud networking services are being implemented to expand applications over various data center sites. The PaaS mainly includes the following features- Operating system Database management system Support Storage Hosting Network access PaaS has provided benefits to software developers, web developers and business firms [5]. The PaaS assists the software and web developers each and every stage of the software development process. They can develop, test and ultimately use the software. The PaaS is also used by businesses who are seeking to develop their own internal software. They can use the ring-fenced development and testing environment for building applications. Following are some benefits of using PaaS to application developers- No infrastructure development needed. Easier for Non experts. Pick and Choose features for flexibility. Security and adaptability. The developer can work together from different locations. IX. REFRENCES 1. Ayadi, I., Simoni, N., and Diaz, G., 2013. Naas: QoS-aware Cloud Networking Services. IEEE 12th International Symposium on Network Computing and Applications. 2. Cole, C., et al., 2013. Higher Order Modulation for Client Optics. IEEE Comm. Mag. pp. 50. 3. Duan, Q., 2011. Modeling and Performance analysis on network Virtualization of composite network-cloud service provisioning. 2011 IEEE World Congress. pp. 548-555. 4. Ferretti, S., et al., 2010. QoS-Aware Clouds. IEEE Cloud Computing (CLOUD) 2010. 5. Francesco, M., and Gianni, F., 2008. An approach to a cloud computing network. IEEE 2008. 6. Gao, J., Pattabhiraman, P., Xiaoying, B., and Tsai, T., 2011. SaaS performance and scalability evaluation in clouds. Service Oriented System Engineering (SOSE), 2011 IEEE 6th International Symposium. pp. 61-71 7. Gringeri, S., et al., 2012. Technical Considerations for Supporting Data Rates Beyond 100 GB/s. IEEE Commun. Mag. 50 (2). pp. S21. 8. Guo, T., et al., 2011. Shared Backup Network Provision for Virtual Network Embedding. IEEE ICC 2011. 9. Xu, M., Cui, L., Wang, H., and Bi, Y., 2009. A multiple QoS constrained Scheduling Strategy of Multiple Workflows for Cloud Computing. 2009 IEEE International Symposium on Parallel and Distributed Processing with Applications. 10. Ahlgren, B., Aranda, A., Chmouil, P., and Queslati, S., 2011. Content, connectivity, and cloud: ingredients for the network of the future. Communications Magazine, IEEE. 49(7). pp. 62-70. 11. Jackson, K., Ramakrisnan, et al., 2010. Performance Analysis of high performance computing application on the amazon web service cloud. Cloud Computing Technology 2010 IEEE Second International Conference. pp.159-168. 12. Jensen, M., Schwenk, J., Gruschka, N, and Iacono, L., 2009. Technical Security Issues in Cloud Computing. IEEE 2009. 13. Kantere, V., Dash, D., Francois, G., Kyriaklopoulou, S., and Ailamaki, A., 2011. Optimal service pricing for a Cloud cache. Knowledge and Data Engineering, IEEE Transactions. 23(9). pp. 1345-1358. 14. Lightner, M., and Widegren, S., n.d.. An Orderly Transition to a Transformed Electricity System. IEEE Transactions on Smart Grid. 1 (1). pp. 3-10. 15. Yu, S., 2010. Achieving Secure, Scalable and Fine-grained Data Access Control in Cloud Computing. INFOCOM, 2010 Proceedings IEEE. pp. 1-9. 16. Milano, F., and Shizhao, Z., 2009. Open Model For Exchanging Power System Data. Power Engineering Society General Meeting, 2009. IEEE 2009. 17. Goodson, G., Wylie, J, Ganger, G., and Reiter, M., 2010. Effective Byzantine-tolerant erasure-coded storage. International Conference on Dependable Systems and Networks, IEEE. pp.1-22. 18. Qinghua, H., et al., 2009. Distributed state estimation with PMU using grid computing. Power & Energy Society General Meeting, 2009. PES ’09. IEEE 2009. 19. Grosse, E., Howie, J., Ransome, J., Reavis, J., and Schmidt, S., 2010. Cloud computing round table. IEEE Security & Privacy. 8(6). pp. 17-23. 20. Shah, M., et al., 2008. Software architecture for delivering power system applications using Software-as-a-Service model. TENCON 2008 – 2008 IEEE Region 10 Conference. 21. Wang, C., Wang, Q., Ren, K., and Lou, W., 2010. Privacy-preserving public auditing for storage security in cloud computing. Proc. Of IEEE INFOCOM’10. 22. Wang, G., and Eugene, T., 2010. The impact of Virtualization on network performance of amazon ec2 data center. INFOCOM, 2010 Proceedings IEEE. pp. 1-9. 23. Yu, H., et al., 2011. Survivable Virtual Infrastructure Mapping in a Federate Computing and Networking System under Single Regional Failures. IEEE GLOBECOM 2010. 24. Bertand, M., 2012. Collaboration interface for network service invocation by Cloud Management Systems. Cloud Networking (CLOUDNET), 2012 IEEE 1st International Conference. 25. Zhou, M., and Shizhao, Z., 2007. Internet, Open Source and Power System Stimulation. Power Engineering Society General Meeting, 2007. IEEE 2007. 26. Zhu, Y., et al., 2006. Algorithms for Assigning Substrate Network Resources to Virtual Network Components. IEEE INFOCOM. Read More