Cloud Computing Theory and Practice – Case Study Example

Case Study 2 Infrastructure as a Service, Platform as a Service and Software as a Service Infrastructure as a service, also known as IaaS, is generally more intricate in comparison to platform as a service (PaaS). While PaaS largely deals with software interfacing, IaaS offers integration that goes deep into the hardware level. PaaS provides a more detailed methodology to current developers’ need for cost-effective applications (Catlett, 2013:43). It also offers the relevant hardware design and software platform required to launch an application commercially without having to manage, maintain, and own all the necessary resources. In comparison to IaaS, PaaS provides comprehensive solutions for the development, creation, piloting, and launching of clients’ application mixed with the connectivity of team members, version control, client web service, software architecture management, and database integration.
Compared to PaaS, IaaS offers tools for networking, hardware, operations and data storage with internet use, so that clients are no longer faced with acquisition and location expenses. SaaS makes use of the internet to offer applications run by third-party vendors and whose platform is accessed on the subscribers’ side. Majority of SaaS applications can be managed directly from web browsers and do not need any installations and downloads, although some need plug-ins (Marinescu, 2013:26). Owing to its online delivery model, SaaS, unlike PaaS and IaaS, eradicates the need to install and use applications on single computers. Using SaaS, it is easy for businesses to align their support and maintenance because all processes (runtime, OSes, applications, servers, data, storage and networking, virtualisation, and middleware) can be run by vendors.
Developers’ Need for Capital
I agree with the proposition that developers with new ideas for new internet services no longer need the large capital outlays in hardware to launch their service or the human expense to operate it. I concur with this notion because web applications have become more dynamic and highly adaptable. This means that they do not need to be hosted on specific hardware as it was once the case. Developers are also creating web applications based on common platforms, meaning that they can be used and manipulated by the common hardware and software platforms (Rountree and Castrillo, 2013:31). A major factor in the reduction of hardware and human investment is greater sharing and collaboration. Technologies like cloud computing have allowed different developers to create and host large sizes of data on the least space possible. Cloud computing enables individuals and organisations to avoid buying expensive hardware and hiring people to operate them. Since they can store their data virtually in common “clouds,” developers can create highly advanced internet services rapidly and manage them easily because the costs are increasingly low.
Obstacles to Cloud Computing
Availability/Business Continuity
Organisations agonise over whether utility computing applications will come with enough availability. This leads some firms to become highly suspicious of cloud computing. Intriguingly, current SaaS offerings have set a high benchmark in this regard (Rountree and Castrillo, 2013:37). For example, Google Search is known for being available to the extent that even the least interruption is noted by major news outlets. Clients expect and demand identical availability for new products and services, a very difficult prospect.
Data Lock-In
Despite software stacks enhancing interoperability between platforms, the storage APIs associated with cloud computing are still basically proprietary, or have at least not been the focus of intense standardisation (Rountree and Castrillo, 2013:42). As a result, clients cannot easily access their data and programs on one portal and run it on others.
Data Confidentiality
In spite of recent developments, security is still one of the most commonly referenced reasons for opposition to cloud computing (Marinescu, 2013:38). Doubtful companies and analysts constantly question the ability of customers to trust that their important data is secure despite being out in the open.
Data Transfer Bottlenecks
Applications keep growing more data-intensive. Assuming that applications could be stretched across the limits of clouds, this may make data transport and storage more complicated (Rountree and Castrillo, 2013:45).
Performance Unpredictability
Although multiple virtual machines are able to share main memory and central processing units (CPUs) remarkably well using cloud computing, sharing of disk I/O and networks is more challenging (Marinescu, 2013:47). Consequently, various EC2 cases differ in their I/O capabilities than in primary memory performance.
Scalable Storage
While the impacts of short-term usage on computation are clear and straightforward, it is less pertinent how to use it on persistent storage (Marinescu, 2013:62).
Bugs in Large-scale Distributed Systems
Eliminating faults in large-scale distributed systems is very difficult. A common feature is that these errors cannot be replicated in smaller sizes, meaning the debugging must happen at scale in the manufacturing data centres (Catlett, 2013:17).
Scaling Quickly
Despite pay-as-you-go being applicable to network bandwidth and storage, which are both used by count bytes, computation is quite different (Catlett, 2013:29). For instance, while Google AppEngine varies automatically depending on load increases and reductions, and clients pay depending on cycles used, AWS charges hourly for the number of times customers occupy, even if their computers are idle.
Reputation Fate Sharing
Bad behaviour by one customer can adversely affect the image of others using similar clouds (Rountree and Castrillo, 2013:23). For example, the number of applications which can be stored can be constrained by spam prevention services’ vetoing of ECS IP addresses.
Software Licensing
Today’s software licenses typically limit the machines on which cloud computing software can run (Catlett, 2013:28). In addition to buying the software, users pay maintenance fee on an annual basis.
Cloud Computing Economics
Cloud computing has no limits on organisations’ specific applications, security and compliance needs and performance regardless of their size. Effective application can offer major savings, a higher degree of stability, and better IT products and services (Catlett, 2013:59). For organisations of all sizes, cloud computing enables cap-ex free computing, low costs, on-demand and flexible scaling, redundancy and resilience, and lower maintenance expenses. It also supports the rapid deployment of projects and supports innovation.
References
Catlett, C. (2013) Cloud computing and big data, Amsterdam, IOS Press.
Marinescu, D. (2013) Cloud computing theory and practice, Boston, Morgan Kaufmann.
Rountree, D. & Castrillo, I. (2013) The basics of cloud computing: understanding the fundamentals of cloud computing in theory and practice, New York, Syngress.