Cloud computing - Cloud - Outsourcing IT

Cloud computing, cloud, outsourcing IT are correlated terms especially when we consider them from functional perspective. Therefore it's difficult to separate them, since the cloud computing was the cornerstone for creation of the cloud, which founds the outsourcing (questionable may be the private cloud). However, the cloud computing, from which it should be started, is a term with interesting history and possibilities. Its history can be reduced to "I had a dream" of informatics whom lived in times when virtualization was developed, data transmission capabilities increased and costs of purchasing technology were less. This enabled thinking about unification of dispersed technical resources and use virtualization tools to create a homogenous, more efficient infrastructures.

"Cloud computing is a technology which combines two hitherto been known methods for data processing - Grid computing and Utility computing. First of them is based on (on - P.M.) so called parallel processing, where a large number of servers are connected to each other and by combining their power we get a far more efficient computing unit. The second conceal term of so called virtualization, thanks to which on one server we can separate some space and i.e. specified bandwidth."[1]

Cloud, which, as it turned out, was a natural consequence of development of cloud computing has evolved at rapid pace. The firsts, which used this tool were Internet services[2], in the case of giants - like eBay or Amazon, which support a large number of countries, have a huge product portfolio and wishing to have their services be adjusted to the needs of their customers as much as it possible - the necessity to use the cloud became obvious. Simple reasons leads to such of decisions, in the first place technical strain on resources at this kind of scale of operations makes investment in their own infrastructure extremely problematic, in the other hand if we assume the possibility of business constant development we should also accept that the need of scalability of resources is necessary - a cloud enables this seamlessly. Ebay: "notes 8 million URL requests every day, this means that as many times every day someone, somewhere in the world enters an address eBay.com into web browser (...). The Number of data that need to be processed by computing machines reaches up to 50 petabytes per day (...) [While - P.M] in the service [Flickr.com - P.M] were published till now nearly 500 million photos and every one of them is displayed in 4 or 5 sizes"[3]. In situation when we deal with a huge amount of processed data to which must be fast and stable access, and downtimes should be minimized, it requires creation of gigantic infrastructure, security system etc. By moving operations to the cloud all mentioned needs are eliminated because then provider of service is responsible for an hardware, access and safety. However, not in all cases, such a solution makes sense. It is not possible in the cloud to export and process huge amounts of data i.eg. geological projects due to the ratio of the amount of data to transmit opportunity links; It is just like trying shedding of the sea through a funnel. In such cases, the mainframe will continue to be used and not replaceable by outsourcing highlights cloud computing.

Cloud is not an homogeneous creation, models and distribution systems vary from each other depending on the concept by which they were created. The most popular models are: private cloud (1), public cloud (2), hybrid cloud (3). 1: this type of model is reduced to the size of its organization, it has a unified security policy, the costs associated with maintaining / buying hardware and licenses, finance staff are debited in full any organization where such a solution is implemented. Implementation relies on virtualization of physical resources, its operation does not imply direct IT outsourcing, but can be classified a single organization that apply private cloud even when it uses on (within a specified range) an external resource provider. 2: this type of model is a cloud en bloc because contact with the service / the service is carried out only via the Internet. Their caretaker is an external supplier, which transfers costs related to infrastructure, security system and staff from the recipient to the service provider. Important for this model is that it can be easily increase or decrease efficiency according to temporary needs. The service is available immediately, and there is no restriction on the nature of conducted activities (there is no reduction according to branches or fields). Although there are good reasons for using a public cloud in each case, it is not always such a solution would be possible. Reason for this is the issue of data security. In the case of private cloud, we can control data, building a security system that prevents a greater extent from being violated by third parties. Although the public cloud provider must comply with EU ordination it should be noted that the law for implementation claims in such of cases like, for example, data security is verified according to frames and legislation of country where service provider has its organization. In cases where organization already holds its own resources and wants to increase their capacity with help comes the virtualization - which also favoured private cloud in that situation. 3: this type of model is a synthesis of 1 and 2, infrastructure and system tools such as applications are divided between the company and the supplier. It result from locating a greater need of scalability a part of activity. Using this type of model organization bears lower costs associated with building their own infrastructure, increases capacity and automates processes where there is a such a need. In cloud computing models have been developed to distribute tools such as applications for companies, which we can use to manage administrative and infrastructure etc. In this case we use browser so there is no need to install the software on each device. It is on subscription fee, which balances costs in contrast to a single purchase of licenses, hardware. We have the ability to personalize graphics and some of the features but we can't expect a complete customizability, still we have to keep in mind that the supplier shall watch over updates, security, speed. Interestingly, a large number of such services (SaaS) allows, if it don't satisfy our needs entirely, to easily use another even in the case of  already-in-use applications. Summarize, all of these elements may entirely consist of the increasing interest in this type of model. The PaaS model is targeted for environments where demand is focused on creating programs, applications - it is a type of platform with customized tools for specific needs and tasks. " The PaaS users receive from suppliers a virtual and fully adjusted to their needs work environment. This kind of model clouds provide inter alia by Microsoft through the platform Azure and Amazon via the Elastic Compute Cloud"[4]. While the IaaS model is targeted for organizations where there is a need to have only technical resources. "The customer buys a specific number of servers, disk space or memory computing. The Clouds user has the option to install and configure the operating system, database systems or final application. In the IaaS model are being sold inter alia dedicated servers that we use in case of necessity of having a good back-hosting[5]". As it's shown by the above illustrated models of clouds distribution can be seen that span covers practically each section of the company from the perspective of infrastructure, ready-made  products or frameworks used to create them.

Praises spread about  automation, cost reduction, speed, simplicity of implementation the clouds will testify in her favour. Opportunities offered and set  trends shown that its should be taken into account in planning investments. The problem facing companies it's not just their burden with buying a server, computers, matrices. Keep in mind that a large pool of costs is not at all mentioned, eg. power consumption in server rooms, system of monitoring (floods, fire, overload, vibrations) employment of security staff, administration and many others creates enormous financial burden. Choosing a cloud a reduction of this burden is possible. However, always remains a "but", which may cause that the company will continue to fight the arduous battle for the revitalization of resources, personnel, etc. because they are worried about the security of their data. Time will show whether the standards to be met by foreign suppliers will fairly implement and actually present. Nevertheless the interest is enormous, which contributes to solving these problems and doubts that are implicated by what should be hoped that the opportunities offered by cloud will not be misapplied, and convinced of its qualities will grow.


[1] P. Szmit, Cloud computing historia, technologia, perspektywy, s. 7. [w]: https://www.web.gov.pl/g2/big/2012_06/ebfa211f1a9f174c7517738f68df2d8b.pdf

[2] vide Ibid, p. 5.

[3] Ibid.

[4] Ibid, p. 11.

[5] Ibid.

This website uses cookies. By continuing to use the site you are agreeing to its use of cookies. More information Privacy Policy I AGREE