• Samuel Kahura Wachira

Evaluation of Current Computing Research on Long Term Data Storage

Abstract

The increasing number of digitized documents and the content explosion that has been experienced in this digital age has necessitated the research into the Long Term Data Storage. There are various methods, modes and approached to this problem. Whichever the solution, it must adhere to these principles: the media or technology should be affordable, flexible, durable and readily accessible and readable by any software. This paper will be examining the current state of available technology for long term data storage, especially on the development of long term storage, the various existing and upcoming hardware, the issues that affect Long Term Storage of Digital data and some possible solutions. In this paper, strategies for preservation of data in specific formats and mediums are also investigated.

1.0 Introduction

Long-term data storage is considered to be one of the biggest topics discussed in the field of Information Technology today. This is attributed to the fact that valuable information faces the threat of becoming unreadable owing to overreliance on paper as a medium of storage. In the modern face the danger of becoming unreadable thus the need to digitize the documents by use of Long Term Digital Preservation (LTDP). Factor et. al. (2009), describes Long Term Digital Preservation as being a set of processes, strategies and tools which can be to store and thereafter access digital data for periods long enough in the wake of changing technologies, formats, hardware, software as well as technical communities.

There has been an increase in the amount of digital information, as well as the aging of existing digital heritage. Well-publicized examples of such losses have resulted into an urgent need to deal with the challenge of long-term preservation of data. It is therefore important for us to understand the components (both hardware and software), the technology, the standards and the research that is being conducted to give us a stable , larger, faster capacity storage system.

We are also able to appreciate the steps that we have made as regards storage size, speed, durability and stability od data storage devices.

1.1 Development of Long Term Digital Storage

A study by Kremser (2012), on long term digital storage revealed that as time goes by, there are many risks that threaten data integrity. Some of such threats include: deterioration and/or obsolescence of storage media, obsolescence of the data format or in general the software needed for running the application for accessing the data such as old operating systems. Such a risk is closely related to hardware architectures.

It follows therefore that in order to ensure long term data preservation one has to overcome the problem of non-prompt memory institutions, including libraries, museums, archives as well as other cultural heritage institutions.

Magnetic tape showed the way to consumer-grade data holding equipment. Since then, storage has pervaded nearly every gadget , electronics, form printers, Home appliances, mobile devices etc. This consumerization is driving the market forces to research more and innovate.

Farley (2009) argues that careful selection of the electronic medium essential in order to deal with the many hazards that result from overreliance digital storage media which are not instantly apparent.

It is evident that data to be archived needs to be retrievable; otherwise the purpose of archiving is negated. In this regard therefore, the use of the Internet could as well eliminate the need to have a single point data storage in the future, if at all the data can be stored and at the same time be retrieved simultaneously on numerous different host computers given that each individual computer upgraded or replaced, the data is therefore retained on other host computers.

2. Technology and Long Term Data Storage

“Technology comprises of discoveries in sciences, product development and improvement in machinery, process, and automation as well as information technology” (Kurnet, 2012). These developments are changing the way that service firms and consumers interact, and are raising a host of research and practice issues relating to the delivery of e-service. Electronic-service is becoming increasingly important not only to determine the success or failure of electronic commerce, but also in the provision of consumers with a superior experience in regard to the interactive flow of information.

Technology has become a leading ‘driving force’ nowadays, in different businesses (Ahmed et. al. (2011). As such it is important to research the investments in technology and their impact in the telecom business. It is particularly essential to assess how technology is reducing the ‘labour intensive activities, reducing service as well as processing cost, increasing service levels, not forgetting improving the productivity and competitiveness of the telecom sector.

1.3 Economics of Long Term Digital Storage

Rosenthal et.al (2012) has some interesting findings as regards the Economics of Long Term Digital Storage. In the paper of the same title, the authors analyse and present facts that support the conclusion that the decrease in cost per bit of storage in this decade, is likely to be much slower or it might eventually stop. “This in turn will make the expenditure commitment implied by a decision to preserve some digital content much bigger and much harder to predict than would be expected on the basis of history”. Rosenthal et.al (2012).

Additionally, Rosenthal et.al (2013). and Goldstein et.al (2010) agree that there will be an increase in the importance of developing accurate and predictive models of storage as well as other preservation costs, as IT budgets become ever smaller.

When organizations recognize the Total Cost of ownership of preserving data for long periods, they migt consider investing in SSD (Solid Stata Drives) which although are initially expensive, will realize a lower TCO than traditional disks, whilw at the same time improve performace of data read and write operations. Rosenthal et.al (2013).

1.3.1 The OAIS Model

According to CCSDS Secretariat (2012), OAIS (Open Archival Information System) is an ISO standard that is intended to be a referential design of digital archives with long term preservation capabilities. Three basic roles are defined in the model: producer, consumer and management.

The OAIS Model (CCSDS et al 2012)

The OAIS model defines functional entities Ingest, Data Management, Archival Storage, Access, Preservation Planning, and Administration.

OAIS is more than a strict specification of referential model. OAIS is a set of vague advice on how to build a digital archive. It identifies some basic actors/roles, describes the data and their flows in the systems. The central notion of OAIS is an information package. There are three types of packages SIP (Submission Information Package), AIP (Archival IP), and DIP (Dissemination IP) depending on the life cycle of the data. The content of those packages slightly differs, but in general there are some descriptive and structural metadata, the manifest describing the content and the data itself.

1.4 Standardization of Formats for Data Archiving

In order for data to be readable in future, it will have to adhere to some form of standards. Hajicek and Studensky (2005) were of the opinion that SGML would eventually become the de-facto standard for data archiving, since it covered most of the criteria for archiving. Standard Generalized Markup Language is considered to be an international standard when defining device-independent, as well as system-independent methods of representing texts in electronic form (Goldfarb, 2010).

However, Hajicek et.al (2005) also suggest that SGML itself does not provide us with a solution for presentation of non-textual data and dynamic nonlinear (interactive) documents.

XML is a lightweight cut-down version of SGML that keeps just enough of the functionality so as to make it useful. It is therefore aimed at making SGML usable easily.

This discussion shows the importance of standardization in the march towards long term data storage.

1.5 Data Storage Media

The selection, storage as well as handling of media is very important when it comes to data storage. This is because we need to understand the various storage media that has been in existence, and is still being developed. Some Storage media has also been overtaken by technology, e.g. floppy disks, resulting in rapid discontinuation of previous formats.

1.5.1 Magnetic Media

Magnetic media makes use of magnetic particles that which are suspended in a non-magnetic film or metallic disks that are then magnetized..

Magnetic media is a flexible, low cost storage medium that has the ability to retain magnetic charges. It requires Clean operating conditions as well as environments to reduce the scope for damage to media and devices.

1.5.2 Optical Media

Optical storage media on the other hand uses laser light to read data from the discs. Optical Storage Media exists in different types such as CD-ROM as well as DVD-ROM (Digital Versatile Disc – Read Only Memory). Ross & Gow (2012),

2.0 Proposed Modes for Long Term Digital Storage

2.1 Disk

The disk industry’s roadmap always predicted a consistent 40% per year improvement with regards to bit density on disk platters, which can be translated to a 40% per year reduction in cost per bit stored. Although in the recent past the industry did not achieve this roadmap target. It is predicted that there is going to be more than a 20%/yr improvement in bit density for the next five years.

2.2 Tape

Tape is considered to be an important medium for long-term storage of large amounts of data. This is attributable to low media, costs, low power consumption as well, its relatively high reliability outweigh its long access times. The recording technology that is used by tape is about 8 years behind disk; however it is on approximately the same cost per bit curve as disk. Thus it is expected that tape’s cost per bit will continue dropping, albeit somewhat more slowly as compared to previously, for perhaps another 8 years. This is likely to increase tape’s cost advantage significantly.

2.3 Solid State Memory

Solid State drives store data via electronic integrated Circuit systems as memory to store data persistently. SSD drived do not have moveable parts, so they do not suffer from the problems of disks i.e Head-Crash, Overheating, and are better at shock resistsnce than their counterparts. Previously, SD Drives were more expensive than disks and were of less storage capacity, but in the last two years, we have seen an incredible leap from 16Gb to 500GB drives.

On a Cost per byte analysis, SSD is still more expensive than traditional disks, but this is envisaged to slowly change as people adopt the new technology and more research and development is carried out.

3.0 Challenges of Storing Electronic Data

Ahmed et. al. (2011), argues that preserving electronically-held data presents a problem. The storage media is unstable and is susceptible to physical abuse. Also, the technological advances that are creating lower TCO drives and discs, is leaving in its wake a lot of obsolete equipment still holding the data. The access of this data becomes a problem because of the scarcity of the equipment that is required to read the data presents problems which mostly seem to be “unsolvable in the face of accelerating technological change”.

Ahmed et al (2011) goes further and concludes that optical discs might offer a compelling solution. However, the breakage and scratching that is usually characterized by storage of optical disks might render his argument null and void.

Data storage on the internet also seems a compelling alternative where companies pay to store and retrieve the data they require, without the headache of managing the storage devices. Amazon, Rackspace, Google and Microsoft are already on this exciting phase of data storage. However, issues of legislation, privacy, security and accessibility of the data is keeping some people away from this solution.

Magnetic Media is cheap, lasts long and is flexible. However, magnetic media are also considered to be vulnerable to corruption via magnetic and electrostatic fields apart from physical and chemical changes that occur during storage. (Russell, 2012).

Good technology should only be applied in data storage if it satisfies some basic principles. Proper technology should ensure that information is stored accurately and retrieved immediately it is needed. All technological advances in automation processing, mass data processing and elimination of administrative challenges fall in this category. Farley (2009).

5.0 Conclusion

The research has shown that there is still a lot of research to be done as concerns Long term Data Storage. Commercial entities are pushing each other to greater heights of research and innovation in the race to produce the cheapest, largest and fastest devices. This race has seen a lot of hardware becoming obsolete, and in essence, not being supported. The question becomes how to read and extract the data in these obsolete storage media.

On the other hand, the introduction of standardized models for long term Data Retention, such as OAIS, will go a long way in helping to achieve the standardization of formats, methods and approaches.

On the economics of storing data in the cloud, Rosenthal et.al (2012, 2013) concluded that it is not yet economical to store data on the cloud, as compared to Disk.

References

Ahmed A, JoAnne Holliday, D. E. Long, Ethan L. Miller, J., and Thomas S (2011). Data management and layout for shingled magnetic recording. IEEE Transactions onMagnetics, 47(10)

CCSDS Secretariat (2012). Reference Model for an Open Archival Information System (OAIS), Recommended Practice, CCSDS 650.0-M-2 (Magenta Book) Issue 2, June 2012

Factor, D. Naor, S. Rabinovici-Cohen, L. Ramati, P. Reshef, J. Satran, D.L (2009) Architecture for Preservation-aware Storage. MSST 2007: 3-15.

Farley, J. (2009). ‘An Introduction to Archival Materials; new media (PRO Preservation Guide series).

Hajicek C, Studensky I, The Long Term Data Storage, FI MU Report Series, July 2005

Kryder, H.M and Chang K. (2009). After Hard Drives-What Comes Next? IEEE Trans. on Magnetics, 45(10), 2009

Neil B. (2012).Digital Preservation Benefit Analysis Tools Project. http://beagrie.com/ krds-i2s2.php, 2012..

Rosenthal, D.S.H., Rosenthal, D.C., Miller, E.L., Adams, I.F., Storer, M.W. & Zadok, E. (2012). The Economics of Long-term Digital Storage. Paper presented at The Memory of the World in the Digital Age Conference, Vancouver, BC. Retrieved from http://www.lockss.org/locksswp/wp-content/uploads/2012/09/unesco2012.pdf

Rosenthal, D.S.H, Vargas D.L (2013) Distributed Digital Preservation in the Cloud,

International Journal of Digital Curation (2013), 8(1), 107–119.

Russell K. (2012). Digital preservation: Ensuring access to digital materials into the future.