A Study On Issues In Cloud Computing

Published Date: 02 Nov 2017

Abstract:

Cloud computing is sharing of resources on a larger scale which is cost effective and location independent. Resources on the cloud can be deployed by the vendor, and used by the client. It also share required software’s and on-demand tools for various IT industries. Usually benefits of the cloud are enormous. The most important one is that the customer don’t need to buy resource from a third party vendor, instead they can use the resource and pay for it as a service. A part from the benefits, cloud got many issues when it comes to security especially on Data theft , Data loss and Privacy. In case of security cloud sometimes suffers a lot. The vendor for cloud must make sure that the customer does not face any problem such as loss of data or data thefting. There is also a possibility where a malicious user can penetrate the cloud by impersonating a legitimate user, there by infecting the entire cloud which effects many customers who sharing the infected cloud. Major problems in cloud computing are Data Integrity, Data Theft, Privacy issues, Infected Application, Data loss, Data location , Security on vendor level and Security on user level. In this paper we are going to these issues.

Keywords: legitimate, enormous, malicious, impersonate.

Introduction: Present industry commonly refers to the cloud as if it was an entity,definable place or recognizable set of parameters. It will be more helpful to understand the cloud as a two way use of computing resources. These resources are available to users on-demand, through the Internet or other network connection. Basing on this cloud model has given rise to three variations.

PUBLIC CLOUD: In this public cloud model, infrastructure is hosted by a third-party provider and is shared among multiple customers. Each customer is charged for computing resources on a metered or pay-as-you-go basis.

PRIVATE CLOUD: in this private cloud model, resources are dedicated to a single enterprise, with infrastructure shared exclusively among the enterprise’s own applications. Private clouds may be hosted by a third party provider , or built within a organization’s own data centers.

HYBRID CLOUD: A hybrid cloud combines both public and private elements on premises and hosted under a common management system. This model has gained great interest among enterprises .Few cloud providers equipped to support a true hybrid cloud. Unless the cloud resides solely in the provider’s domain, incompatible delivery platforms and inflexible management systems are often unable to accommodate legacy systems and configurations that every enterprise data center contends with.

Architectures:

There are many different architecture in the cloud computing . This paper is concentrating on two service layers, the bottom resource layers and the upper service layers. The lower layers consist of virtual resource layer and the physical resource layer while the upper layer consist of software as a service(SaaS), platform as a service (PaaS) and infrastructure as a service(IaaS) . The infrastructures as a service in the architecture supplies computing resources and storage resource for users. In case of any particular service constraints, IaaS provides an intermediate platform to run arbitrary operating systems and software. Platform as a service is in the middle part of the cloud layer which gives users better performance a more personalized hardware and software services and a lot of infrastructure module such as infrastructure module, distributed data module, the user registration module, billing module etc. These modules could be used as a service (SaaS). SaaS provides application which is closest to the user‘s services and allows deploying the software in a network environment so that the software can be run under a multi-user platform.

Traditional Security computer and the network intrusions or attacks are now more possible to the cloud environment. However most of the service providers have proved beyond reasonable doubt their services are been protected without any fear like this to occur to any company‘s data. Most companies have decided to keep less sensitive data on the cloud and reserve the sensitive data to them. It has been proven as well by Jericho Forum that it is much easier to lock down information entrusted into the third party than when information is to be managed in-house, because it is easier to enforce security via contract with online services than via the internal control.

Availability security, relies Critically on the applications and being available, one of the most prominent incidents of such security is the Gmail: one day outage in mid October 2008, Amazon S3: over seven hours downtime on July 20 2008 and Flexi scale: 18 hours outage on October 31, 2008. The providers are confident that their uptime is comparable better when considering the in housed data centers in their companies. In some cases, there is more risk of one single failure when company subscribes a single point cloud provider. If anything happens with that provider, then, it will affect the data and whatever infrastructure within the providers platform.

cloud.jpg

The legal implication being held by third party is somehow complicated and it is not well understood by many people. There is always lack of control and confidence when a third party has held a secret. This has made a lot of private companies to build a cloud of their own in order to avoid such chaos taking place. In order to control data in the cloud to getting access to everyone there is a means to shift from protecting data from the outside (system and application which use the data) to protecting data from within. We call this approach of data and information protecting itself as information-centric (hht://news.cnet.com/2100-1030_3_6102793.html). This self protecting requires the effort of intelligence. Data needs to be encrypted and packaged with the usage policy in place.

Security Issues faced by Cloud computing :

images.jpg

When it comes to Security, cloud really suffers a lot. The vendor for Cloud must make sure that the customer does not face any problem such as loss of data or data theft. There is also a possibility where a malicious user can penetrate the cloud by impersonating a legitimate user, there by infecting the entire cloud thus affecting many customers who are sharing the infected cloud. Some of the problem which is faced by the Cloud computing,

1. Data Integrity 2. Data Theft 3. Privacy issues 4. Infected Application 5.Data loss 6.Data Location 7. Security on Vendor level 8. Security on user level

Data Integrity:

When a data is on a cloud anyone from any location can access those data’s from the cloud. Cloud does not differentiate between a sensitive data from a common data thus enabling anyone to access those sensitive data’s. Thus there is a lack of data integrity in cloud computing .This fear can only be mitigated by trust, and trust is earned by providers based on their track record of performance. Select a provider with a long history as a market leader. Look for the provider to share not just product descriptions, but details about how it operates. Finally, look for management tools that provide high levels of visibility into your cloud environment. For example, users of IBM’s cloud services can utilize the company’s robust Tivoli management system to manage all their IBM cloud and data center services, whether on-premises or hosted, private or public.

Data Theft:

Most of the cloud Vendors instead of acquiring a server tries to lease a server from other service providers because they are cost affective and flexible for operation. The customer doesn’t know about those things, there is a high possibility that the data can be stolen from the external server by a malicious user.

Privacy Issues:

The Vendor must make sure that the Customer Personal information is well secured from other operators. As most of the servers are external, the vendor should make sure who is accessing the data and who is maintaining the server thus enabling the vendor to protect the customer’s personal information.

Infected Application:

Vendor should have the complete access to the server for monitoring and maintenance, thus preventing any malicious user from uploading any infected application onto the Cloud which will severely affect the customer. Enterprises planning to host workloads that are latency- sensitive should discuss their expectations for end-to-end performance with their provider. While few (if any) public cloud offers include end-to-end application performance service level agreements, some providers will custom-configure a solution that meets specific workload needs, usually combining a private cloud with a private network configuration. For each workload, be sure to consider the provider’s standard service assurances, as well as its history of meeting those commitments.

Data Loss:

Data loss is a very serious problem in Cloud computing. If the vendor closes due to financial or legal problems there will be a loss of data for the customers. The customers won’t be able to access those data’s because data is no more available for the customer as the vendor shut down.Enterprises should expect that their sensitive cloud workloads will have at least the same level of protection as their sensitive on-premises workloads; but for less sensitive workloads, they should avoid paying for excessive security. First, understand what base security solutions e.g., firewalls your provider has built into its cloud architecture, and what assurances it is willing to offer in the case of breaches or loss. Is it sufficient for the workloads you’re planning to put into the cloud? Second, apply a strong user authentication scheme associated with your cloud Workloads. Some cloud service providers, including IBM, offer optional security services that allow you to tailor the security solution to the workload.

Data Location:

When it comes to location of the data nothing is transparent even the customer don’t know where his own data’s are located. The Vendor does not reveal where all the data’s are stored. The Data’s won’t even be in the same country of the Customer, it might be located anywhere in the world.

Security on Vendor level:

Vendor should make sure that the server is well secured from all the external threats it may come across. A Cloud is good only when there is a good security provided by the vendor to the customers.

Security on User level:

Even though the vendor has provided a good security layer for the customer, the customer should make sure that because of its own action, there shouldn’t be any loss of data or tampering of data for other users who are using the same Cloud.

PROTECTING THE CLOUD A Secure cloud is always a reliable source of information thus protecting the cloud is a very important task for security professionals who are in charge of the cloud.

Some of the ways by which a cloud can be protected are Protection of data, making sure data is available for the customers, delivering high performance for the Customers, using Intrusion Detection System on Cloud to monitor any malicious activities, to make sure the application used by the customer is safe to use, Vendors must provide a support system for the customer, customer should be able to recover any loss of data in the cloud.

Most important of them all is that, there should be a good degree of encryption provided by the vendor to the customer that only the customer should be able to access the data and not the malicious User.

Conclusion:

Both the Vendor and the customer should make sure that the cloud is safe from all the external threats, thus there will be a mutual understanding between the customer and the vendor when it comes to the security on Cloud.

The two important steps to take before entering into any cloud computing agreement is to identify the risks described above to determine the client’s comfort level with respect to each and begin the discussion of risk balancing early in the negotiation. Lengthy, unsuccessful negotiations can be avoided if each side is clear as to their "deal-breakers" with respect to these risks up front.