Cloud Computing – the Past, Present and Future
10 years ago when I started my PhD my research topic was High Performance Computing. I finished my PhD in Grid Computing and I am now teaching Cloud Computing systems. In my (actually not that long) professional life I have experienced three paradigm shifts, all of them promising cost efficient sharing of computing infrastructures, ease-of-use, non-trivial Qualities of Services (QoS) and accounting/metering systems as easy as those used for accounting and metering of electricity. Well, the first two paradigms failed to deliver what they promised. It is more than justifiable to ask what about the third one, what about the Clouds? How sustainable is Cloud technology? Or it is just another buzzword, a placeholder until the fourth paradigm appears? In the following I will try to answer the most common questions that I am faced with: as a Cloud researcher and Cloud user.
How can I use the Clouds? Well, most of us are already using Clouds – without even being aware of it. And most of us are very surprised by that fact – since we do not pay to use Clouds – at least not in the common sense. However, in Clouds (as in many other ICT areas) the users pay with their privacy and their data. For example, many SaaS providers like Facebook, Gmail, etc. use our data, either directly or they derive aggregated user data to deliver targeted advertising and product placement. So, the challenging question for the future is: what is the “Cloud currency”: our privacy, our data – and are we aware of that?
I am a (SME) company with computing resources that are (sometimes idle). How can I provide my resources to the public use? Many small and medium European enterprises that are actually doing other business are suitable to act as providers by offering their idle resources to the public. However, they have difficulties in entering the Cloud markets due to lack of appropriate easy to use, configurable and versatile infrastructures and tools. Existing infrastructures and tools are usually not tailored to the needs of small/medium companies and do not consider issues like a reduction of costs, adherence to energy consumption and load fragmentation or plug and play solutions for provider defined cost models. There is need for versatile tools with customizable cost models incorporating explicit management of non-functional requirements like energy efficiency, that operate on autonomous infrastructures reflecting cross-organizational structures.
How can I find suitable Cloud offers?
Currently, there is a huge gap between the current Cloud computing offers and the actual demand of enterprises and communities. While a wealth of work has been accomplished to suit the technological development of Clouds, there has yet been very little work done in the area of the market mechanisms that support them. As identified by many scientists the existence of appropriate Cloud markets where customers have the choice between comparable set of offers, without provider lock-in and with the low entrance barriers to the market will decide whether Cloud computing will take roots as a new state for the art technology for implementing large scale distributed systems or not. However, the establishment of open Cloud markets is faced with many challenges like appropriate definitions of the virtual products traded, interoperability between platforms, and sophisticated market management. Current Cloud markets are either portals where big players (Google, Amazon) offer their products leading consumers to the provider-lock-in or they trade an unorganized plethora of Cloud products, are faced with fluctuating trade base, and rely on inefficient static market management mechanisms. There is need for self-adaptive markets that can automatically optimize their core properties to changing environmental conditions, fluctuating trader participation or malfunctions in the infrastructure.
What technological impact will Cloud Computing have?
The effects of Cloud computing cannot be fully foreseen today, but ubiquitous access to seamlessly unlimited computing power could change the way we are living. The penetration of Cloud technologies to our daily life could change the way we work, for example in the area of data management. Today, huge quantities of data are continuously generated, stored and maybe never again accessed. Yet, our knowledge is hidden in the data we produce, and has to be discovered by intelligent, versatile, and efficient data analysis. Cloud computing could deliver a novel approach for a holistic, socially enhanced and evolutionary knowledge generation process.
What about socio-economic impact that Cloud Computing might have?
An example of the socio-economic impact of Clouds is in our education systems, as they do not usually keep pace with technological development. We will face, with the new role of IT departments in companies, a demand for new profiles and curricula of computer scientists, resulting in the need for a mixed curricula education in subjects like law and economics as well, of course, as computer science.