Technical configurations and business models
Taking these definitions and characteristics as the point of departure, we can infer two key underlying objective elements that define the cloud:
i) The technical configuration
ii) The business model
The technical configuration represents, firstly, the kind of hosting platform, and the different essential characteristics that allow hosting of websites/applications on virtual servers, which pull computing resources from extensive underlying networks of physical web servers shared by multiple users. The business model includes the service model, the type of cloud deployment, and other contractual agreements between the provider, client, and any third party agencies. This includes how the server space is allocated to different users based on the fees paid. Different combinations of the technical configuration and business models can be said to underlie the ‘cloud infrastructure’. For example, the state information technology (IT) department of a ministry can run their own data centre and a private cloud, where they offer services for free or at a fee to the health department. Alternatively, there can be a contract where the state gets a third party to host their application on the public cloud, using Linode, at a monthly fee. And, there can be various different combinations of the technical configuration and the business model, making it important for us to be clear about their meaning when the term ‘cloud’ is used.
While the cloud is seen to be cost-effective and elastic, Vaughan-Nichols (2015) argues that there is nothing like a free lunch, as currency fluctuations (in the USD, Euro, or Norwegian Krone) may suddenly hike up real prices. Recently, the three big public cloud services—Google Cloud Platform, Amazon Web Services, and Microsoft Azure—raised their prices. Azure did so by 13 per cent in the Eurozone and 26 per cent in Australia, as also did Microsoft by 26 per cent. Business models are governed by forces of capitalism, and subject to the necessary influences which come with it. This raises the need to examine alternatives of keeping data in-house where there can be more control over the costs, or use hybrid models of the cloud to hedge risks of global currency fluctuations.
The cloud infrastructure is typically made available to users through the ‘service’ model which provides for process storage, networks, and fundamental computing resources. It is made operational through an ‘outsourcing’ model as followed in the general IT software and services area, but with a key difference. Traditional software outsourcing was done from the rich (such as United States and United Kingdom) to the not so rich countries (like India, Vietnam, and China) primarily for cost advantages, and because necessary human resources were not available in-country. Starting from the 1980s, this outsourcing industry has flourished, today becoming a thriving and profit-making one, particularly in countries like India, providing gainful employment to large numbers of people. Interestingly, cloud outsourcing works the other way round, where many LMIC users and governments access cloud hosting service in the rich countries, because they are able to provide the cloud infrastructure at more cost-effective prices and more reliable services than what typically would be provided for by companies from LMICs.
In public health informatics this seems an attractive proposition. One could, for example, use online computing to replace hundreds of offline facility-level installations with one central, online server. This should potentially enable integration and render superfluous to a large extent the problem of end users having to do technical maintenance of local hardware and software. However, as we saw in Chapter 4, the terms at which information is produced, disseminated, and consumed contributes to how it is valued and used. New problems may emerge through the use of central online servers or the use of someone else’s infrastructure (such as global companies) through rental agreements. For example, control could shift to large global companies and their affiliates who run the cloud infrastructure, giving them, potentially, control over the data. These issues are part of a spectrum of concerns related to control over information and overlap debates about net neutrality, open data, development, data sovereignty, and similar such issues. In-between these broader questions over the emerging political economy of information, we also discuss the operational aspects of the cloud and big data, and how LMICs are trying alternative models to operationalize them, and what some of the empirical experiences have been.
In summary, we argue that the cloud must be viewed in terms of a confluence of business model and technology, which are, to a certain extent, autonomous with different driving forces but also very much in dialectic relation to one another with respect to nature (power, cooling, green technology, etc.), labour (who does work for whom, and when and who gains value from the work), everyday life (e.g. the actual health consequences to the non-professional ‘beneficiary’), and more. Further, the global geo-political-strategic importance of information, surveillance, security, big data, and various other non-economic factors of production are also at play.
While the focus of this chapter is primarily on the cloud, in line with our Expanded PHI approach, we argue that discussions on the cloud are incomplete without discussing big data (what is stored in the cloud) and data analytics capabilities (ability to pull data out of the cloud). These have implications on the institutions and people involved, and their inter-relations. The identified issues around the cloud are explored in the context of these inter-connections.