In this post we analyse economic cost and legal particulars of local hosting of Data.
Local hosting of data is perhaps required to reduce data transfer costs, assures access to important data (particularly in case of internet outages) and ensures security of data. In the current commercial context these are the basic advantages of local level hosting over cloud hosting. However, when idea of local level hosting is looked at from the perspective of reducing cost of access to consumers and the same is qualified by the idea of ‘data of community interest’, following infrastructural and policy issues arise:
- What is the cost involved in hosting data at the local level? And how to reduce cost of local hosting?
- How to determine the data of community interest?
- What are the other issues that may arise of local hosting policy? (Fundamentally against the principles of net neutrality)
In case of the proposed setup (proliferation of broadband by Wi-Fi network), basic cost will include server’s acquisition cost, setting up the server, housing of server, uninterrupted electricity supply for the server and hence electricity backup, maintenance cost etc. Hence, fundamental concerns in this case is that whether this cost of setting up a server and subsequent maintenance of the same will be less than the combined cost of data transfer and cloud (external) storage of locally relevant and locally generated data. In order to determine the same, we have to firstly look at advantages and disadvantages of the two. Thereafter we can determine commercial viability of local hosting or determine measures to make local hosting commercially viable.
Local level hosting
- Data is secure.
- Run any application without any boundaries since full control is localised.
- Faster access of files, backups and printers (distance is less).
- Local access on the actual hardware in case of failure.
- No need for Internet access in order to access user’s files.
- Issue of internet outage largely decreases.
- Initial capital is very high in relation with cloud services.
- Maintenance costs might be high in case there are several servers.
- Require constant electricity supply.
- IT support and will be required for the maintenance of the hardware and software of the server. (requirement of manpower)
- Moving server to remote inaccessible locations.
- Limited space for data storage.
Cloud hosting or hosting on the internet
- Stable monthly costs without any initial capital.
- Limitless expansion of your servers based on your needs.
- Savings on maintenance costs.
- Access to your data from anywhere without any additional cost.
- Pay as you grow schemes which correspond to cost savings in the long run.
- Not feasible to use any application that you desire.
- Your data is stored in the cloud provider’s data centre.
- Very difficult to change cloud service providers.
- Accessing to your data might be slow since it is based on your Internet connection.
Commercially speaking, setting up of local servers requires large sum of capital and maintenance of such large number of servers, at geographically far-off places will also require large investments. This large initial investment overrides the long term benefit that may arise out of fast and unrestricted access to the internet, security of data, reduced cost of access etc.
Under such circumstances, where initial cost is high but the service pays over time, incentives have to be provided at the initial stage. These incentives will encourage initial cost bearing and over time local level hosting will pay for itself because of the reduced costs and subsequently service will become cheaper over time. Initial incentives can be provided in following three manners:
- Government funded: Government may subsidise installation of servers at the local level. This one time cost, which is indeed substantial, can be transferred directly to the private players or a public sector unit like BSNL may be asked to undertake the oversight of the setting up business. Once local servers have been set-up, management and up-keep of the same can be transferred to private firms. These firms can in turn be subjected to regular audits in order to determine the quality of maintenance. Tax rebates for setting up local servers to those operators who undertake Wi-Fi operations may also work as an alternative to subsidising the setting up of local servers. However, fundamental problem with this model is that government cannot continue to subsidizing maintenance and up-keep of local servers forever. This would be an inefficient approach. This might very well require some model of cost distribution to the local user base or content developers.
- Transferring cost to the local user base: Cost of setting up and maintain local servers might very well be transferred upon the local user base. The best method to do this is to increase the cost of access to data stored over the local server. This method would only work if access to locally stored information is differentiated from access to information accessible from the rest of the internet. This would mean that locally stored information will have to be either accessible at a faster speed, will have to be more secure etc. or there will have to be specific information that is only accessible by using local server. The bottom-line is that locally stored information and access to the same has to be differentiated from the rest of the information in order to make this option commercially viable.
- Transferring the cost to content providers: Issues of net neutrality arose because, “Internet Service Providers (hereinafter ISPs) attempted to collect fees from large content providers, arguing that such fees would allow them to upgrade their hardware to accommodate the growing demand and better serve the end user. Some ISPs proposed another way to raise revenue, charging content providers for a “fast lane” that would give priority to their content and ensure faster delivery to the end user.” There is perhaps no harm in using a similar approach to finance local storage of data. A proportion of storage space can be exclusively used for storage of content of the content developer and the remaining proportion can be used for storing data of community interest. However, fundamental problem with this approach is that it involves basic net neutrality issues and TRAI has already declared any form of preferential data transfer illegal.
As per the researcher, a hybrid of above three will be the best solution for encouraging setting up of localized servers and subsequent local storage of data. Setting up of servers in rural areas where consumers cannot bare cost can be subsidised by the government. On the other hand, cost of such localization can be transferred directly to the consumers where they are financially capable of baring such cost. This includes increasing the cost of access to localised content on Airports, restaurants etc., perhaps the places where luxury tax is imposed. Lastly, places where substantial number of people access certain form of content, such cost can be transferred to the content developers. This could perhaps be done in market places, taxi stands etc.
Once infrastructure is created to host data at the local level, next big challenge is to determine whether what content or data is of community interest. The basic issue faced while determining the same is that different people in a community have different priorities and interests. Consequently determining community includes becomes difficult. This difficulty is further depend by the fact that no comprehensive legal definition of community interest exists. Perhaps looking at public interest may be of some help in determining community interest. Public interest may be defined as “1. The welfare or well-being of the general public; commonwealth. 2. Appeal or relevance to the general populace: a news story of public interest.” Considering this definition, following two methods may be adopted in order to determine community interest:
- Public Trust Doctrine: The Public Trust Doctrine primarily rests on the principle that certain resources have such a great importance to the people as a whole that it would be wholly unjustified to make them a subject of private ownership. The said resources being a gift of nature, they should be made freely available to everyone irrespective of the status in life. The doctrine enjoins upon the Government to protect the resources for the enjoyment of the general public rather than to permit their use for private ownership or commercial purposes.
This idea, along with the fact that explicit arrangement has been that private owner access to the publicly owned spectrum and rights of way necessary to exploit the technology is exchanged for public access and speech rights, ensures that public trust doctrine be applied to the proliferation of broadband by Wi-Fi. This is due to the fact that information being stored locally is simply an extension of larger internet and hence is a public utility. Further, a faster and secure access is being given to certain information because of it being important for public or in the interest of the community.
Hence, keeping the above two principles in mind, the power to determine the nature of information available on local servers should be vested with the government or any other body accountable and answerable to the government. Government may subsequently determine information of public importance and depending on the same determine the nature of information that is to be stored locally.
- Local community usage patterns: Local community may be allowed to determine the nature of information stored on the local server. This may be done by identifying the local data usage pattern. Hence, information accessed most frequently by the community may be identified and stored locally. Such information in turn may be collected by identifying the internet usage pattern of local community. This will not only ensure cheap access to such information but will also ensure reliability of access considering the unreliable state of wire based broadband proliferation infrastructure. However, the basic drawback of this approach is that community may not store data important for it in the time of crisis. Hence, this method should only be adopted after due deliberation and behaviour tracing algorithms should be balanced for variables that are relevant for communities, but generally not considered by them.
- Hybrid approach: This might very well be the best approach. This approach requires the government to determine the usage of a proportion of locally available space of the server and the remaining space may be used for storage of information depending on the local community’s usage pattern.
Despite of all these infrastructure and policy considerations, saving and granting faster and cheaper access to selected content is against the core principles of net neutrality. Telecom Regulatory Authority of India’s ruling dated February 8, 2016, rules against this approach. However, regulation 6may be used by the authority to relax norms in this case. Further, if above mentioned guidelines with respect to determining nature of data of community interest are considered, relaxations can be granted considering that it is the government and the local community that discriminates data, and not the ISPs or the content developers.
 Efficiency of government subsidies can be seen in various cases. Few examples: http://www.forbes.com/sites/gregsatell/2013/07/02/4-government-programs-that-drive-innovation/#6a1b951b64f2
 P. Carter and J. Mayers, Division and Distribution of Community interest in defined benefit pensions (Available at http://lawschool.unm.edu/nmlr/volumes/18/1/04_carter_division.pdf)
 Available at http://www.legalserviceindia.com/articles/ptdoc.htm; M.C. Mehta v Kamal Nath.
 Internet was held to be a public utility in U.S.A’s federal communication’s ruling with respect to net neutrality. Available at http://www.nytimes.com/2015/02/27/technology/net-neutrality-fcc-vote-internet-utility.html?_r=0; Telecom Regulatory Authority of India held something on similar grounds.
 This may include local weather forecast, information related to agriculture etc.
 W. Aiello et al, Analysis of Communities of Interest in Data Networks, AT&T Labs-Research.
 Researcher recommends this approach.
 PROHIBITION OF DISCRIMINATORY TARIFFS FOR DATA SERVICES REGULATIONS, 2016.