The Cloud Database Model - Research Paper Example

? CLOUD DATABASE Table of Contents 3 Introduction and Background of the Research 4 Critical Analysis of Cloud Database 6 Advantages of CloudData 10 Disadvantages of Cloud Data 10 Conclusion and Recommendation 13 References 14 Abstract In the 21st century, the biggest asset of the company is data and proper organization and backup of data requires was one of the issues that led to the evolution of database management. Companies throughout the world are globalizing and diversifying their operations across nations. From each and every transaction records, customer information, employee records, knowledge management system of organization, etc. to all business related and competitor information constitutes huge volumes of data. Organizing and maintaining such enormous amount of data in only papers is almost impossible. If back up is not maintained properly, the business might lose critical information which could lead to losses. Thus, the migration of existing and current data to cloud based platform will require well defined architecture as well as strategies. The cloud database model should have specific objectives in line with policies of the organization. The model should also include sufficient security, encryption, and control over information which will be transferred. The biggest challenge lies in maintaining the balance between speed of data transfer, cost of data transfer, data accuracy and consistency. If all these aspects of cloud database management are not ensured then the consequences could be data loss, wrong location data placement or data alteration during transfer process. Even though cloud based computing platforms are becoming popular, especially for e-business and MNCs, due to cost efficiency, there are issues related to data safety, speed and security. The objective of this paper is to critically analyze the cloud database by pointing out disadvantages of cloud data. Introduction and Background of the Research The cloud computing technology has been receiving significant attention in the market lately. For many sectors and companies in various industries, cloud is not just an umbrella for marketing but a consolidated technology that offers grid computing, SAAS (software as a service), distributed computing, and so on. There are many companies that offer cloud services including Amazon RDS and Microsoft SQL Azure. A cloud database is basically a database management system that runs on cloud computing platform (Curino, pp.1-2). Some of the cloud computing platforms are Go Grid, Amazon EC2, and Rack space. (Source: Curino and et.al, n.d, p.2) The main vision on the basis of which cloud computing was developed is based on certain key features that cloud database is expected to perform. Some of the important features that was kept in mind while creating cloud data platform were data storage, processing, software installation, software delivery from local servers and desktops across various networks. The key concept was that the target users must get access to remote data from anywhere at any time without requiring to carry a hard drive or other storage device. The concept was created to free big corporations from generating own power and capital investment. Thus, the most important candidate of cloud computing is the database management applications. The main reason behind this is that large corporations spend huge sum of money to safely manage enterprise data which involves upfront investment in both hardware and patented software services. The advantage that the different enterprises would get from cloud computing is that they could reduce their annual data management cost significant with the help cloud database. By using cloud data the companies will not have to install, own and maintain dedicated expensive software packages. The reason is that cloud computing service providers or vendors provide their customers with virtual machines or platforms from where the customer will be able to install and manage software services without having to own and store a version in their local hard drive. Thus, their cost and resources is saved and can be better allocated elsewhere to generate revenues. Such concepts may sound tempting but the objective of this study is to critically analyze cloud database and explore its disadvantages and advantages of deploying database system in cloud. The research critically analyzes the typical properties of popular commercially accessible cloud computing database platforms that have the potential to affect the choice of managing data and applications that can be deployed in cloud technology. Critical Analysis of Cloud Database The aim of this section is to research various articles on data management in cloud computing platform and explore which database management application should be best suited for managing cloud data. This section will also critically examine the various disadvantages in data management models of cloud data. In order to accomplish the research aims, it is important to first discuss the most typical cloud characteristics. (Source: Green Host IT, 2013) Power Elasticity – The computing power of cloud is elastic only if the workload can be distributed across parallel networks. Elasticity refers to the data independency when conditions or networks are changed. For instance, when the demand for the product and services of the company is at peak, or may be during the exponential expansion stage of a social networking company the volume of data increases exponentially with possibility of infinite data management requirements. With the help of data elasticity feature of cloud database, the growing companies will be able to manage their data properly without additional cost. While in the earlier days, it took the company many days to install and manage additional hardware and software resources to organize sudden peaks in data volumes. With the help of cloud computing database this problem is eliminated to an extent. But the power of elasticity cannot be increased magically and in real life situation it primarily depends on memory of work station, CPU’s power as well as local capacity. The additional resources are required to allocate respectively to additional servers in order to accomplish surplus data management activities. For instance, the Amazon’s EC2 (Elastic Computer Cloud) can create private virtual instances of servers by offloading some works in the new servers. But such activity has limitation that the server will only be able to create large number of instances if the new servers are running in parallel with older instances. Without this basic criterion, the cloud platform will only be able to handle limited data and thus additional servers will not be much of help (Abadi, 2009, p.2). Data Storage at Un-trusted Host – The relationship between a buyer and a seller is that of trust and similarly many businesses has been successful today because their customers trust these companies. It is thus impervious that the host company violates the privacy of customers and let third party access data without the consent of customer. But due to biasness and prejudice in many customers, there is always a shadow of doubt in their mind that their confidential data might be transmitted to unsafe hands. Such nervousness pulls back customer demand and affects company’s profit margins. It cannot be denied that transfer of data across public or private networks has potential security risk as the data can be interrupted in the intermediate path and hacked by unwanted parties. In order to avoid such unwanted intrusions the cloud computing platforms must ensure proper and appropriate precautions while the data is transferred from one network to another. The cloud database management system often gives an impression to user that the data is stored and delivered at remote location. However, in reality, the data may be physically located in a particular host country such as the United States. The host country has set of guiding rules and regulations. For instance, according to the US Patriot Act the US government will be able to access data of third party without the consent of person or company who is hosting the data. This gives limited control to user regarding where the data will be stored and who will be able to access data. Hence, from the above discussion regarding data storage or cloud data it can be said that the user will have limited choice but to assume the most pessimistic scenario. There will be limited options left for data security and control unless proper encryption key logic is implemented. Again there is a serious limitation of cloud data regarding encryption that is, if the encryption key is not located at host the confidential data can be accessed by unwanted and unauthorized parties without even the knowledge of actual customer (Abadi, 2009, pp.2-3). Data Replication – One of the most important aspects of cloud data is its durability and availability. Any loss or unavailability of data could damage the goodwill as well as the bottom line of business. The durability and availability of data at any place or time depends on data replication. This is because if the data is located at a single site, any major hard drive crash may make data unrecoverable. To avoid such circumstances, cloud computing companies spreads data across many data centers which are geographically separated. These data centers are indentified as data zones where original data is replicated and stored. So in case the facility of a particular region fails, the company will still be able to recover data from other available data zones. Thus, it is important for the customer to understand the concept of replication scheme and if the data is replicated in same region or zone, data failure chances increases. It was also found in the research that the principles of ACID (Atomicity, Consistency, Integrity and Durability) were hard to maintain especially when data was to be transferred from one geographic location to another. Thus, data replication concept for easy availability of data may become a hurdle and serious limitation of cloud database (Abadi, 2009, pp.2-3). Risk of Storing Intermediate Transactional Data – The transactional databases of cloud data consists of all relevant business critical process and information that completes the operational needs of the organization. The data might also contain sensitive customer details such as address, email, telephone contact, or even credit card number. Thus, when the data will be transferred from one site to another, any serious security breach will violate customer privacy and might lead to severe consequences including law suits. Hence, the security during transfer of information is very important and violation of privacy is totally unacceptable. Ability to Operate with Encrypted Data – As discussed earlier, sensitivity of data is very essential in cloud database and that the data must be encrypted before it is uploaded on to the cloud platform. Thus, in order to prevent unwanted access to private customer data, any application that would be running on cloud must not have the permission to decrypt data without the consent of customer. Another limitation to note here is that transfer of entire relational database via cloud, at a time, is bandwidth intensive. Capability of Running in Heterogeneous Environment – The performance of cloud computing is not consistent due to many reasons such as hardware compatibility, non-virtual resources, slow speed in some networks due to uneven traffic, disproportionate latency which may be region specific and so on. If appropriate measures are not taken then the ability of cloud computing in different settings and locations is affected (Abadi, 2009, p.7). Advantages of Cloud Data Cloud databases offers noteworthy advantages over traditional database management systems with the help of features such as automatic failover, increased accessibility, fast automated recovery, broader scalability, minimum investment, minimum maintenance, and cost efficiency. Disadvantages of Cloud Data It was also found during the research that cloud database also has potential drawbacks especially in areas related privacy, security, data consistency, remote access, potential data loss, or inability to access critical data. As discussed earlier, security is the biggest concern in case of cloud database. With the help of cloud platform a company will be able to transfer private data of individuals and corporate. It is the responsibility of the cloud computing platform service providers to ensure data protection, proper management and consistent data retention. Thus, it can be said that cloud service provider’s reliability is very critical. Sometimes the sensitivity of data is so important that the very existence of the company may be jeopardized in case of negligence by the service provider. Similarly, the privacy and accessibility of data is also very important issue in cloud database. The company or individual will have no choice but to rely on their service providers for data protection and prevention from unauthorized access. There is a general panic between the users regarding password leak or data theft which is often triggered from various stories. Such stories might make the users feel uncomfortable while granting permission to access their data to third party. Another major disadvantage of cloud database is implied dependency on the service provider also known as vendor lock-in. The vendor lock-in concept leaves the users with limited choices since once any user has registered with a service provider then it is almost impossible to switch to another service provider. Thus, the portability of service provider is very cumbersome as the huge volume of data has to be transferred again from stage one in case the user decides to change the service provider. Technical issues with cloud database are not unusual when it comes to very large organization. The smaller companies generally prefer subcontracting technical maintenance part to an established and reputed IT company. But it must be kept in mind that any system, no matter how powerful it is, might face technical dysfunctions time to time. The cloud service will definitely be impacted due to down time from slow connection. This is especially because the whole cloud database setup and model is dependent on internet access (Rogers, Papaemmanouil, and Cetintemel, pp.1-2). The internet connection may be slow due to data traffic or network latency which in turn will affect the performance of the data transfer. In general case, if the connection is slow the user may get frustrated due to time taken for transfer of file. In some cases, if the volume of transaction is very large the system may revert the entire transaction which although maintains data consistency but definitely wastes time. In case of larger corporations, latency in network connection would ultimately lead to business loss in the form of reduced profitability and declined performance. Also, the cloud technology has been found to take several minutes when it creates a new instance from the default server. (Source: Rogers and et.al, n.d, p-2) Since all applications and services in cloud computing runs on third party or remote virtual environment, the users as well as the companies will have limited control over execution and functioning of software and hardware of cloud database. This feature restricts the users from running the application in local server. As discussed earlier, such characteristic of cloud database will make user data vulnerable for malicious attacks. It is least to say that nothing on the internet is completely secure since it is a public network and due to this reason even the biggest players suffer from serious security breaches and attacks. Thus, interdependency of system has the potential to compromise privacy of data. Conclusion and Recommendation During the research it was found that database systems deployed on cloud faces many new challenges such as dealing with power of elasticity, scalability of operations depending on the size of organization and the volume of data, autonomous control and security issues related to transfer of data and its remote access. These challenges can make the cloud database system highly prone to faults (Agrawal, pp.5-12). References Abadi, D. J., 2009. Data Management in Cloud: Limitation and Opportunities. [Pdf]. Available at: http://www.cs.yale.edu/homes/dna/papers/abadi-cloud-ieee09.pdf. [Accessed on June 22, 2013]. Agrawal, D. and et.al, No date. Database Scalability, Elasticity, and Autonomy in the Cloud. [Pdf]. Available at: http://www.cs.ucsb.edu/~aelmore/papers/dasfaa.pdf. [Accessed on June 22, 2013]. Curino, C. and et.al, No date. Relational Cloud: A Database-as-a-Service for the Cloud. [Pdf]. Available at: http://people.csail.mit.edu/nickolai/papers/curino-relcloud.pdf. [Accessed on June 22, 2013]. Rogers, J., Papaemmanouil, O., and Cetintemel, U., No date. A Generic Auto-Provisioning Framework for Cloud Databases. [Pdf]. Available at: http://cs.brown.edu/~jennie/research/provisioning.pdf. [Accessed on June 22, 2013]. Green Host IT, 2009. Pros and Cons of Cloud Web Hosting. [Online]. Available at: http://www.greenhostit.com/company/green-blog/93-web-hosting/244-pros-cons-cloud-web-hosting. [Accessed on June 24, 2013]. Read More