Policy for Supporting Grid Education and Training

Status of This Document

This document provides information to the Grid community on policies for supporting Grid Education and Training. It does not define any standards or technical recommendations. Distribution is unlimited. This is a draft version of the document, it has not yet been submitted for public comment, but is still being developed within the ET-CG group.

Copyright Notice

Copyright © Open Grid Forum 2008. All Rights Reserved.

Abstract

The development of e-infrastructure, of which grid computing is a fundamental element, will have major economic and social benefits. Online and financial businesses already successfully use grid computing technologies. New research methods and technologies generate large data sets that need to be shared in order to ensure continued social and scientific research and innovation. Grid computing systems can provide an environment for coping with these large data sets and for sharing data across regions. An investment in educating people in grid computing, then, is an investment that will strengthen our economies and societies. In order to deliver e-Infrastructure education internationally, we must develop a policy framework that will ensure shared responsibility and equivalent training in the field. This document introduces the current challenges for grid and e-Science education and training and presents opportunities and existing structures for education and training, as a starting point for further work. It then proposes strategies and policies to provide a supportive framework for e-Infrastructure education and training.

Contents

1. Introduction – why invest in grid education?

2. Definitions

3. Challenges for grid education

3.1 Curricula and textbook development

3.2 General Expertise in e-Science

3.3 Teaching and the fluidity of the technological landscape

3.4 Disparate educational policies

3.5 Sharing training infrastructure

3.6 IPR and sharing

3.7 Training-specific challenges and requirements

3.8 Impact of standards on education and training

4. pportunities and existing structures for education and training

5. Suggested strategies and policies

5.1 Strategies

5.2 Policies

6. Future work

7. Contributors

8. IPR statement

9. Disclaimer

10. Full copyright notice

11. References

12. Appendices

1. Introduction – why invest in e-Infrastructure education and training?

The NSF presents a definition of ‘cyberinfrastructure’ which can be used to describe the various components of e-Infrastructure: “computing systems, data, information resources, networking, digitally enabled sensors, instruments, virtual organizations, and observatories, along with an interoperable suite of software services and tools. This technology is complemented by the interdisciplinary teams of professionals that are responsible for its development, deployment and its use in transformative approaches to scientific and engineering discovery and learning.” [1] It is precisely these teams of professionals that are the focus of the OGF Education and Training Community Group (ET-CG). Investments in e-Infrastructure require adequate investments in e-Infrastructure education in order to allow for the full development and utilization of these technologies.

We can, therefore, identify four primary reasons why it is vital to develop policies to support e-Infrastructure education and training:

1. a skills and knowledge shortage in business, government, academia and society
2. optimisation of the use of e-Infrastructure
3. benefits for industry and academia
4. the transition to knowledge-based economies.

First, we are facing a wide-ranging crisis due to a shortage of e-Infrastructure related skills and knowledge in business, government, academia and society. A 2007 review of IT skills and careers in the UK revealed that skills shortages and skills gaps still plague the field of computer science, and this has knock-on effects in other subject areas and sectors. Skills required to use virtualisation technology, for instance, are sorely lacking to the point that more than half of UK businesses cannot take advantage of this technology. But the crisis is not only felt in the UK. In 2005, IT companies pointed to a European-wide “skills crisis as a shortage of computer graduates and a retiring technical workforce threaten to bite IT departments by 2006.” In the United States, similar realities were identified at the close of 2007: “there is a distinct shortage of certain IT skills, and that shortage seems to be growing.” Concerned computer scientists have particularly pinpointed a lack of expertise in grid computing, explaining that, “Grid may be the liberal arts of computing. It requires knowledge about many IT disciplines, a flexible management approach and acceptance of new ideas. But resumes boasting grid-specific skills and accomplishments remain rare. Grid is not widely taught, and IT workers with hands-on experience in this young field are tough to find.” Education in e-Infrastructure more generally is deficient in its current state, as evidenced by the distinct deficits in skills and knowledge noted above. This crisis cuts across regions and sectors, as e-Infrastructure computing technology proves to be a ubiquitous enabler. If the crisis is addressed, we will find ourselves in a win, win, win situation, in which students gain employability, employers gain skilled staff and educators gain a market.

The second reason to develop education and training policy relates to the first. e-Infrastructure technologies such as grid computing involve the potential risk of poor return on investment if measures to support usage of the infrastructures are not put in place. A compelling example, which applies to any research infrastructure, is that it takes years of training to get the best out of facilities. The complexity, novelty and changing nature of e-Infrastructure means that there is a high risk of under-utilisation, or non-optimal exploitation without adequate investment in education and training. The investment in e-Infrastructure to date has provided a pervasive and dependable platform on which a relatively small proportion of experts can demonstrate the high value of the research and innovation it enables. Today’s challenge is to strengthen this platform so that the realisation of these benefits of e-Infrastructure become routine, that is, any researcher in any discipline routinely uses the resources e-Infrastructure provides as fluently as an artist uses a brush or an engineer uses differential equations. This requires two concurrent and coordinated advances:

1. The educational progress identified in this document, and
2. The steady improvement in the facilities, tools and ease of use of the pervasive e-Infrastructure.

At present, the second branch of this strategic requirement is limited by the lack of sufficient skills across a sufficiently broad spectrum of society and academic disciplines to deliver the advances.

Both industry and academia benefit from e-Infrastructure, or grid computing, outputs or applications, another key motivation for developing policies to progress grid education. Use of e-Infrastructure has already become integral to finance and online businesses, a primary reason for their economic success. In finance, grid computing can solve problems associated with large and complex computations. Data centres at online companies such as Google and Amazon use forms of grid computing to manage the vast number of searches requested by users on a daily basis worldwide. Advances in scientific and other knowledge as well as new technologies have generated vast amounts of data that require proper management. Phil Wadler, Professor of Theoretical Computer Science at University of Edinburgh, observes that, "computing has become a fundamental tool in all research disciplines, which often proceed by compiling and managing large databases and/or exploiting computer models and simulations (a topic sometimes called e-Science)." Today’s research into social and scientific issues and problems is not only more data intensive, but has become increasingly more collaborative, which often involves the international sharing of data. Education and training in the use of e-Infrastructure prepares students to use grid computing and other systems and these systems facilitate better management of data and collaboration.

Finally, countries around the world are transitioning to knowledge-based economies, which rely on the education of citizens in the latest ICT and research methods (etc.). OECD and World Bank country studies have confirmed an obvious correlation between investment in education and quality of life and GDP. There are economic benefits to educating citizens and particularly in preparing them, through education, for the current social context in which we see evidence of the use of computing technologies across academic disciplines and generally in our daily lives. Basic ICT infrastructures now exist in a majority of universities worldwide. The EC, for instance, has recognised ICT as key to a knowledge-based economy and social cohesion, and so it must have a place in education and training. Individuals can make the best judgements and make contributions to the knowledge-based economy if they are equipped with the proper skills to exploit existing and rapidly developing technologies. e-Infrastructure is one such technology which can provide the tools to allow countries to “become better at producing knowledge through research, diffusing it through education and applying it through innovation”ii (integrating the knowledge triangle), in order to successfully compete in the global knowledge-based economy.

It is clear that greater investment must be made in e-Infrastructure education so that a skilled workforce exists to use and further develop e-Infrastructure technologies throughout the world. Without education and training that targets both students in computer science, those individuals who need in-depth operational knowledge of e-Infrastructure, and students in other disciplines, who must know how to use e-Infrastructure to enhance their research capabilities, countries around the world will flounder in their attempts to become players in the knowledge-based economy. As a means of confronting and correcting the skills and knowledge shortage apparent in e-Science and grid computing technologyiii, the OGF Report argues for further investment in e-Infrastructure education and training; it envisions the embedding of education and training into normal academic training throughout the world. The Report presents a list of motivations that justify this vision, after introducing relevant vocabulary. It provides a clear picture of the current state of grid and e-Science education in particular (with broader focus developed after further research and input), highlighting challenges, those areas that need improvement and development, and opportunities, any existing methods and tools that can be used and expanded upon. The Report concludes by setting out strategies and policies that delineate the vision for continued coordinated international growth of e-Infrastructure education, thus allowing for the full exploitation of e-Infrastructure technologies.

We specifically need to:

1. Invest in education in appropriate computational thinking or digital-systems judgement in every scientific, medical, engineering and humanities first degree so that a culture is developed and graduating students are equipped to contribute to the knowledge economy with an appreciation of the potential of e-Infrastructure and rich information sources and well prepared to make competent ethical and socio-economic judgements about their use.
2. Invest in education of specialists via undergraduate courses and Masters courses to develop a critical mass of experts who will innovate both in the provision and exploitation of e-Infrastructures and e-Science methods.
3. Invest in Doctoral and Postdoctoral training programmes that develop intellectual and business leaders and educational leaders who will take forward the development of international research capacity in this field.

By harmonising and collaborating internationally, each country will benefit, both from economies in the cost of the required innovation in educational provision and in the mobility of the resulting skilled citizens. The harmonisation also leads to a community of experts and leaders who are better equipped for trans-national cooperation in research, innovation and business.

2 Definitions

e-Infrastructure – the term is used to denote the digital equipment, software, services, tools, portals, deployments, operational teams, support services and training that provide data, communication and computational services to researchers. An e-Infrastructure is usually multi-purpose and has to be a sustained dependable facility so that researchers can plan to use it for the duration of their work. [4]

e-Science – the invention and application of computer-enabled methods to achieve new, better, faster or more efficient research in any discipline. It draws on advances in computing science, computation and digital communications. [5]

t-Infrastructure – e-Infrastructure adapted to the needs of education, trainers and students. Shared t-Infrastructure would be usable by students and teachers internationally, providing easy access to educational exercises running on e-Infrastructure.

Definitions of other terms related to e-Infrastructure education can be found on Appendix A.

3. Challenges for grid education

We can find evidence of gaps in current e-Science education, as well as training, that could pose problems for both students and educators teaching the use or provision of this e-Infrastructure. We use the term e-Science to include grid computing, e-Infrastructures provisions and the exploitation of the computational and data resources they make available. Grid education and training is only one element of the total requirement for education in this field. Further work is planned to develop a broader agenda—contributions and volunteers are welcome. Certain tools or structures are missing in e-Science education and this could hold back development attempts. Understanding these challenges is important when considering what strategy and policy recommendations to propose. The following challenges have been identified:

3.1 Curricula and textbook development

Key challenges concerning curricula involve the need for concerted coordinated work on its development as well as determining various modes for delivery of curricula. Not enough time has been spent developing and defining curricula for grid and e-Science education. The skills and knowledge developed needs to be attractive to industry and academic sectors, since students will be drawn to courses if they are generally assured employment after completion. More time spent on curricula can lead to clarification of the “what” and “how” of teaching grid and e-Science education as well as the drafting of a framework for curricula that can be used internationally.

The ACM produces curricula guides for computer science courses and these can be used as a reference.vi The ICEAGE (OGF-ETTF) Curricula Development Workshop, planned for February 2008, will result in useful collaborations in this area which can also be referenced by the OGF ET-CG. Accreditation bodies such as the BCS and UK Engineering Council, and ACM, could play an important role by certifying courses so that students completing these courses are attractive to industry.

Multiple modes of delivering distributed computing education would be required not only to address the issue of fluidity of the technological landscape (highlighted in a subsequent section) but in order for that education to have wider appeal and relevance and thus greater uptake. Different target audiences would require the presentation of different principles, concepts, and examples, so that the mode of delivery and curriculum are geared towards that audience. Flexible refresher courses could update students on new technologies and summer schools could appeal to academics who would not have time to commit to a Masters course in grid computing.

Grids, web services and other forms of distributed computing allow the pooling of resources, the integration of models and the management and analysis of large data collections. To equip students with the ability to use models appropriately and to have good judgement about the validity and interpretation of results they need experience with models appropriate to their discipline. These may be numerical, stochastic, Baysian, process, statistical or logical models. Additionally, different disciplines have different tools, such as Matlab, that are used for accessing models, organising parameter sweeps and analysing results. The academic curriculum should give the students relevant experience, preferably using examples related to their discipline and academic maturity, of choosing models, planning their use, conducting in silico experiments and interpreting results.

Many disciplines depend on increasing volumes of shared data in public or proprietary data repositories. An archetypal example arises in earth systems disciplines which are concerned with predicting climate change and mitigating its impact. Examples of the scale and complexity of this data can be seen in the European INSPIRE Project.vii The curriculum has to teach students how to find and understand the data relevant to a problem in their field. They need to be able to assess the fidelity and temporal validity of such data, to conduct analyses using that data and interpret the results. The curriculum has to contain relevant examples and be supported by the resources (student-accessible data, computation and tools) that will enable students to develop the relevant understanding and skills.

In many subjects, digital data is collected via a variety of instruments: telescopes, satellites, sensor networks, observational buoys, medical images, social surveys, etc. In socio-economic, political, epidemiological and ethnographic research, the data may be produced as a side-effect of people’s daily activities. Still other data is generated by collaborating communities subscribing to compendia of observations and annotations. Students require an appreciation of the data collection processes and the ways in which data may be post-processed to generate derivative information, to normalise and standardise to deal with equipment and observing variation and so on. The topics taught must again be relevant to the given discipline and develop judgement as to the interpretation of such data.

More advanced students, e.g. those engaged in forecasting the course of exceptional environmental events (floods, hurricanes tornados, eruptions, earthquakes, tsunamis, etc.) require understanding of the challenges of coupling observation and modelling, and of meeting time constraints in delivering results. This may lead on to all of the issues that arise in planning and coordinating emergency response. Data mining is widely used in some disciplines. Students in these disciplines require an understanding of the forms of data mining and the interpretation of the results they produce.

All students need to develop a professional understanding of the ethics of information systems- policy may one day depend on the quality of their advice. They should have an understanding of privacy issues, encryption techniques and security methods. Here again practical and valid examples relevant to the discipline and academic maturity of the students is necessary. It would be good if in the longer term this could build on a core of general knowledge and developed judgement that could be assumed by the e-Science courses.

Consider also the following breakdown of student category and educational needs, identified at the 2nd ICEAGE Forumviii:

• Computer scientists and software engineers—theoretical foundations of distributed computation and insights into engineering trade-offs and current implementation strategies.
• Application developers and users—functional and pragmatic presentation of capabilities, an understanding of performance and cost trade-offs and illustrations tuned to their disciplines.
• System engineers and managers—criteria to assess and select technologies, need to understand operational trade-offs and failure modes, and need to be able to undertake resource planning.

More time spent on curricula can lead to progress in professionalising a new category of engineers specialising in grid computing, to establish professional practices after refining curricula to meet the needs of various types of students. At the 2nd ICEAGE Forum, this need for professionalisation was raised after discussions concerning how to improve on current systems unreliability and failings. [9]

e-Science textbooks

There currently is a lack of adequate textbooks to support curricula. e-Science educators face the challenge of writing good text books, as do educators in other fields, which require clarity and conciseness so that students can grasp complex ideas and concepts. It takes time to know how to teach distributed computing well “as a whole”. You need to know what to teach (what to leave out) and how to teach it (considering method, structure/organisation of material). There is a still greater challenge if you set out to equip students in a cohesive group of disciplines how to take best advantage of e-Infrastructure.

One way to generate textbooks would be to set up a fund to pay for selected leaders in the field to devote time to writing (one year, for instance). Another option would involve the pooling of information on specific sites, sharing this information and debating about what and how to teach, coming to consensus and developing (the outline of) a textbook from this, which can be used internationally (translated). Cooperation on the creation of this textbook would lead to improved resources for teaching (and more efficient development of these resources). A strategy would need to be developed to determine how to go about this and in turn, policy would need to be developed regarding pooled information and its use in textbooks. The SURA Grid Technology Cookbook has recently been made available online and could provide a guide in terms of content for future grid computing textbooks, but also in terms of the collaborative efforts involved in its creation. [10]

It is recommended that incentives be developed, e.g. a competition, in conjunction with established editors and publishers, to develop textbooks that serve and help to define the agreed educational goals and curricula.

Ultimately, the normal commercial processes leading to established and progressively improved text books will probably take over the field, but this depends on developing a market of sufficient size. The initial steps described above are needed to build such a market.

3.2 General Expertise in e-Science: grid computing and computer science

We can identify a lack of “general” experts in the field of e-Science, and a shortage of experienced teachers. Development of education would involve the sharing of material, as expertise in certain areas of e-Science is scattered among individuals. The challenge would be to create a new approach to managing and sharing teaching materials due to this lack of general experts, in order to advance academic and research communities.

It is necessary to prime and stimulate an incremental international growth in e-Science educational capability. This has already started in countries throughout the world, partly due to the effects of the Information Society Digital Infrastructures programmes [11]. It requires a positive feedback loop of the following form:

1. Research on infrastructure R&D generates experts with knowledge of e-Science
2. Some of those experts’ time is then invested in developing curricula, courses and material and in educating a cohort of students.
3. Some of those students enter step 1 with greatly increased skills and knowledge compared with their forerunners and in increased numbers.

Initially step 2 is achieved mainly in doctoral and post-doctoral programmes. To increase the step change in skills, knowledge and capabilities this must now move into the undergraduate programmes.

Computer scientists contributing to the development of e-Infrastructure education will most often be specialists in a particular technology within their field, which can be problematic when attempting to expand education beyond that aspect of computer science or when teaching how methods may be used in a particular discipline. But, computer science need not provide e-Science education across all disciplines. It can, however, provide other disciplines with the basic tools necessary to incorporate grid education into their academic departments, to become a force for sharing materials and allowing access to experts.

3.3 Teaching and the fluidity of the technological landscape

It is difficult to keep up with rapid change in the computing world. Grid technology and associated standards are constantly evolving with new recommendations and software from standards bodies and solution providers. [12] This means that educators have a daunting task, as do students attempting to learn ever-changing material. Grid computing can provide the solution by strengthening collaborations and cooperative networks which can result in better understandings of these changes and rapid international response, leading to advancements across disciplines and an overall increase in competitiveness. An opportunity arises to develop policies and institutions to facilitate fast and fair exchange.

3.4 Disparate educational policies: harmonisation and security

Harmonisation

Pertinent educational policies that already exist in universities and within countries (at national level) are disparate. For example, university grid access policies for students differ from country to country and even within countries; currently in the UK, postgraduates can have access to the National Grid Service (NGS) but project, campus and regional grids can often have a variety of student access policies and this is problematic. There is a need for harmonisation of these policies so that grid computing is introduced (with ease) more broadly within most disciplines. Students and teachers need to be able to reuse skills and experience as they move around the world. There is a need for policy harmonisation or mechanisms to support interoperation, since grid computing is generally international. Grid education can be promoted and use of grid computing can be increased through harmonisation of these education policies, for the benefit of users and providers.

Some students will require practical and specific skills, such as the description and submission of computational jobs, the management and movement of files and the coding of programs to execute in and exploit a grid context. Here the widespread adoption of relevant standards, including in the taught material, is an obvious step towards harmonization. In the examples just given, the OGF standards, JSDL, GSM, GridFTP and SAGA, would probably be the basis for consistent treatment, leading to skill (as well as code) mobility.

Security

Following on from the challenge to harmonise education policies is the challenge of security. Security issues arise as a result of the sharing of resources across institutions and state boundaries, leading to access and use problems. [14] For example, universities issue identity and authority for students to work with their facilities. When students and staff use multi-institution or multi-country facilities some risks of misbehaviour and choice of authority occur. But complex authorisation can inhibit engagement.

In order to move towards policy harmonisation, the conditions of use that would need to be placed on students, home institutions and visited organisations (this division may not be applicable, depending on how grid access and use is determined, but it provides an example of possible tiers of responsibility) and the providers/operators of grid computing services, as well as technical requirements, would have to be clearly defined and communicated. The eduroam infrastructure use policies (including the European eduroam confederation policy) and technical specifications can provide starting points for future work on such requirements and development of harmonised international e-Infrastructure and grid education and training policies. [15]

Students need to be allowed secure and clearly-defined access to and use of resources (what they are allowed to do must be clearly understood) through authorisation structures as they learn and develop knowledge and skills.

3.5 Sharing training infrastructure

The term t-Infrastructure is used to denote the infrastructure that is needed to enable the educational goals to be met, particularly to develop understanding and experience through practical experience. In a sense, it is the e-Science analogue of laboratories in biology. In practice, the t-Infrastructure is the computing equipment, digital communications, software, data and support staff needed to teach a course. The OGF ET-CG has begun to clarify issues surrounding t-Infrastructure provision in the Training Infrastructure Document, which details European experiences with training platforms such as Gilda and Genius and provides world-wide examples including the Open Science Grid and summer school infrastructures. [16]

As indicated in the discussion of curricula above, there are many topics to be taught, and their presentation has to be adapted to the discipline(s) and maturity of the students. To give the students good practical experience requires much investment to develop or acquire the relevant t-Infrastructure. This is illustrated by a number of examples:

1. Experience of a parameter sweep using a computational model. The software incorporating the model needs to be written, licensed or purchased. This can be best accomplished by pooled efforts across institutions. The data used by the model needs to be set up. This may require selection and simplification to make the task tractable for students. The parameter space to be explored needs to be chosen by the educators for similar reasons. The computational facilities to execute the model runs and collect the results for each student must be provided. This is demanding as (a) the entire cohort will submit their jobs at approximately the same time, and (b) the students require a response within a reasonable time and a low rate of failures or learning is impaired. As classes run at different times in different places, there is a good opportunity to take advantage of pooled resources.
2. Experience of data analysis. Let us say the students are given access to a set of predictions of a hurricane’s path and the census and property data of a relevant region and asked to identify areas where the risk times cost is high so that they receive priority. Collecting example data of predicted hurricane tracks is probably relatively straightforward, though a single request to the hurricane centre may be much preferable to many requests to the centre from many educators. However, setting up the census and property data is a much more complex task. It requires negotiation over how much information may be presented. It requires transformation to hide the actual data while still presenting a sensible geographic and social situation. It requires adaptation to show all the educational examples but tractability in the expected time for the expected category of students. The advantage of doing this work once, sharing the cost and re-using it in many institutions and countries is self-evident.
3. Experience of interpreting medical images. As digital scanning methods (e.g. MRI and digital x-ray) increase it is important to educate medical students in their use. The current volumes of data involved can be substantial, as can the computation to render images according to requested viewing parameters. A pooled resource can have several advantages: (a) it shares the collection, cataloguing, anonymisation, ethics negotiation and privacy costs, (b) because it can draw on data from thousands of centres it can have a far more complete collection of rare diseases and rare presentations for a particular imaging technology, (c) because it draws on non-local populations, accidental recognition is very unlikely, and (d) the larger collection may support better atlases and epidemiology.
4. Experience of working in a collaborative multinational and multidisciplinary team. Many research programmes, engineering projects and policy support activities depend today on effective work in such distributed teams supported by the best Computer Supported Collaborative Working, shared computing and telepresence methods. In order that students can be prepared to work in such contexts, they need to undertake projects in their curriculum that simulate relevant aspects of such collaborative working. Setting this up and supporting it require multi-state collaborative action.

3.6 IPR and sharing

A further challenge relating to sharing and trust models involves Intellectual Property Rights (IPR). A framework for sharing in terms of IPR needs to be in place, but so far no models have been widely accepted at international level.xvii In the European context, the 2001 EU Copyright Directive (Directive 2001/29/EC) is an attempt at standardising, or harmonising, copyright law among Member States, keeping in mind certain modern requirements of the information society, and as such it relates to educational materials that would be shared in the case of e-Science (etc).xviii Considering a wider (international) view, the Berne Convention is well-established and addresses the issue of copyright, as does TRIPs, within the World Trade Organisation (WTO) agreements.xix The World Intellectual Property Organisation (WIPO) also provides frameworks for IPR that might be relevant. But the challenges arising for e-Science and the sharing involved in use of e-Infrastructures are relatively new and still in the process of being unravelled and addressed. This issue is being tackled within the OGF ET-CG.xx At present, ICEAGE and EGEE repositories provide (contained) educational materials that can be safely used due to such rights issues having been addressed; rather than copyright, deposit agreements and creative commons licences could provide a model to apply in e-Science education. [21]

3.7 Training-specific challenges and requirements

Training can be distinguished from education in that training is a targeted short-term process to develop specific skills in a certain technical area, whereas education can be seen as an institutionalised long-term process using conceptual models and resulting in development of a culture (but these are by no means discrete categorisations). In order to increase training opportunities in e-Infrastructures, and particularly in grid computing, certain challenges must be addressed, some of which mirror challenges introduced in discussion of education:

• For instance, lack of teachers with appropriate expertise and the problems associated with teaching in the midst of technological change arise in both the areas of education and training. Developing an internationally-recognised certification process which provides teachers with quality training (and credibility) that includes periodic updating of knowledge would be a reasonable response to this challenge.
• The content of training courses on international level, as well as methods of delivery, are currently different, as they are in educational courses, but in the case of training this is often the result of vendor variety (so that each vendor provides training on their product and each product requires unique vendor-specific methods of operation). Definitions of key terms, for instance “security” and “job”, may differ depending on the vendor, based on differences in product.
• Cooperation on development of shared t-Infrastructure would be beneficial in the training arena.

Despite these similarities and overlapping challenges, certain training-specific challenges and requirements can be identified:

• To define the structure of training certifications, considering skills required at each level - work has already been done within the OGF ET-CG to suggest types of certificates, based on skill sets.xxii Three certificates have been proposed: certified grid technician (CGT), certified grid professional (CGP) and certified grid architect (CGA). To obtain the CGT certification, the trainee must complete a base technician module and one specialisation module; the focus is on practical rather than conceptual skills. The CGP would obtain a certificate after completing a base engineer module (more in-depth than the CGT base module), more than one specialisation module and after developing both practical and conceptual skills. And finally, the proposed CGA is trained to have a high-level view of the grid.
• To convince vendors (industry players) to participate in developing a general training process.

3.8 Impact of standards on education and training

As remarked above, much of the e-Infrastructure and specific tools in use vary from site to site and in many cases are also evolving rapidly. This variation and the rate of change increases the cost of preparing and presenting courses, reduces skill mobility and detracts from the amortisation of costs through shared t-Infrastructure.

Ineluctably as some of the education goes hand in hand with research, it is at the frontier and must endure rapid change as understanding, methods and technology develops. However, for the majority of the education neither the variety nor the rate of change is necessary.

It is important that the education and training community work closely with the standards development organisations to encourage the development and uptake of relevant standards. For example, the EU education and training community should then work in concert with technology providers, e-Infrastructure providers and educational institutions to encourage and accelerate the adoption of relevant standards. Just as the units used in a Physics course work anywhere in Europe so should the terms and methods taught in an e-Science course. And, in a wider context, the context that concerns the OGF ET-CG, these terms and methods should be relevant at international level as well.

4. Opportunities and existing structures for education and training

It is important to understand the existing state of grid and e-Science education in order to know what options are out there for educational planners and how to proceed. Research up to now has focused on identifying existing tools and infrastructures within the European context in particular (resulting in the e-IRG ETTF Report), while we are not as familiar with other world regions. In order to provide a more comprehensive view of the international state of grid and e-Science education, in order to explore opportunities, we invite you to provide us with information on your region.

The following identified tools and infrastructure, as mentioned, derive primarily from the EU context, but certainly they reflect what is occurring in many world regions in relation to e-Infrastructure education:

4.1 Existing educational machinery – curricula, t-Infrastructure and security

A number of EU Member States provide Masters courses and summer schools on grid education. Currently, there are Masters courses available in grid computing and related areas throughout the EU. But, there appear to be few coordinated efforts across universities to work together on provisions for the Masters courses. Other (undergraduate) courses and summer schools are run by countries including Greece, Portugal, Germany, Italy, Estonia, Finland, and Hungary. ICEAGE has offered summer schools through the ISSGC series. [23]

A list of university postgraduate courses, summer schools and online courses has been compiled on the OGF wiki and contains examples from around the world, but the list is far from comprehensive and requires updating. [24]

EU Member States have not yet worked together to create a coherent infrastructure, so that there are no shared security networks and IPR (beyond OGF) and curricula are created on an ad hoc basis, without backing from accreditation bodies. Masters courses are aimed at research output (producing researchers) when they could also be aimed at industry through accreditation. Member States could develop a shared t-Infrastructure and shared security and IPR frameworks, as could other regions throughout the world. They could ensure that courses are certified by accreditation (industry and professional) bodies.

4.2 NGIs and the EGI – providing infrastructure for education and training

Building infrastructures is expensive, so coordinating by engaging with regional or national grid system providers already operating in different member states throughout the EU would minimise costs. Coordination that allows sharing of knowledge is also beneficial. Most universities do not have access to all experts in the field, so expert knowledge sharing among institutions would increase the EU’s overall competitiveness in research and innovation. Coordination can lead to standardisation of core material and attainment criteria for education internationally, so that mobility is facilitated. Development of an international infrastructure would advance the sharing of curricula, qualifications and teaching methods.

Existing National Grid Initiatives (NGIs) and the European Grid Initiative (EGI) could provide foundational infrastructure for grid education in the EU. There are developing NGIs in 37 European countries which could in principle provide infrastructure for grid education.xxv As a single national point of contact for local institutions in each Member State, the NGI could connect all fields involved in grid computing and e-Science, providing the following services: easily available and accessible t-Infrastructure for classroom exercises and teaching, identity management and security and tools/techniques for setting up “grid in a box” systems on demand. Such an infrastructure can provide a model for other regions throughout the world and also lead to development of linked regional infrastructures that together create an international infrastructure.

The EGI, currently in its design phase, will help to integrate the NGIs and provide coverage where no NGIs exist (also stimulating development of NGIs in these Member States). The EGI should help in the harmonisation of e-Infrastructure education across Member States through coordination of NGI services such as authentication and security. European e-Infrastructure integration, as well as integrations in other world regions, also has to consider HPC and the PRACE design study.

4.3 Embedded e-Infrastructure in national educational operations, plans and policies

There are already examples of the embedding of e-Infrastructure into national education policies in EU Member States, particularly involving security. In Greece, for instance, students receive their student card, email and grid access upon registration, as part of the existing educational security model.xxvi We need to identify whether there are models that would allow this to happen elsewhere.

When considering opportunities for educators, we have to keep two models in mind: one based on harmonisation and one that allows grid education to develop in an organic fashion, which it is currently doing. There are trade-offs. Advantages to harmonisation are skills transfer, mobility, credit transfer, integration, cost savings and shared curriculum development. Advantages to an organic process are diversity, cross fertilisation which can lead to innovation, meeting national and discipline requirements faster and flexibility to better respond to a rapidly changing domain. We should consider both models when formulating strategy and policy recommendations.

5. Suggested strategies and policies

This OGF information document sets out options to increase engagement with e-Infrastructure technology, and distributed computing in particular, on an international scale. The document has reviewed the current state of grid and e-Science education, presenting related challenges and opportunities. The suggested strategies and policies listed below can support international development of e-Infrastructure education.

5.1 Strategies

Curricula development – Encourage and invest in the interdisciplinary and collaborative development of new grid computing and e-Science modules at departmental, institutional and national levels, and provide means for coordination in terms of curricula:

1. The e-IRG ETTF Report proposes establishing a committee/body of leading educators across disciplines to expedite the creation of the curricula goals and principal topics, launched and supported by major conferences highlighting educational priorities and opportunities in the field.
2. Continue meetings in international contexts, such as that in Brussels and at OGF 22 and 23, to develop understanding of educational goals and curricula.
3. Continue to build a repository of shared experiences and practice in e-Science education (see Appendix B for a list of Masters and other courses offered in each EU Member State).

Develop a means to pool information, cooperate and provide standards of use for information to produce textbooks and other teaching material for grid education. Options for production of adequate textbooks include:

1. Establishing specific websites and other relevant fora where information for textbook content can be pooled, shared and debated about.
2. Setting up a fund to pay for a selected leader in the field to devote a block of time to writing a textbook.
3. Developing incentives such as competitions, in conjunction with editors and publishers, to produce textbooks which follow agreed educational goals and curricula.

Refer to the SURA Grid Technology Cookbook and network with contributors regarding content and collaborations. See Appendix C for an expanded list of curricula and textbook development resources.

Investigate changes to education already occurring as a result of emerging ICT and changes that could be made.

Certification – Encourage certification of courses by professional accreditation bodies, build on current harmonisation and cooperation. At the 21st OGF, E&T sessions included discussion of establishing the Grid Professional Institute (GPI). This institute could in fact be formed by existing bodies such as the BCS (for Europe) and the ACM (for the Americas) for the purpose of managing certification internationally.

Promote the sharing of resources – Investigate shared security models, for t-Infrastructure, relating to existing procedures to move towards standardisation by embedding e-Infrastructure in a similar manner in the national education policies of all Member States. It was suggested at the 2nd ICEAGE Forum that a task force should be set up to assess existing tools, their ease of use and suitability, including security issues. Best practice could be determined after plumbing current models.xxvii

Address challenges concerning the sharing of materials, considering IPR and repository provisions.

Develop relationships – Look at national and international e-Infrastructure to support education to determine what relationships to develop. Providing stronger links between the BCS and the ACM, in relation to certification would be beneficial. Consider relationships between the EGI, EGEE and Open Science Grid (OSG). Further input is invited here.

5.2 Policies

We can identify the need for two kinds of policy in order to establish a framework for shared responsibility and equivalent educational training:

• Policy for providers of education. These would be common rules to address issues arising from the sharing of ideas, software and computing.
• Policy for teachers and students. These would be common rules to address issues arising from equipment use (so students do not crash systems) including conditions of use and mobility and the need for access (to allow continuity of work, for instance, with PhD students).

The OGF ET-CG therefore recommends the development of policies on the following issues:

1. Recommendations as to the level of investment necessary (nationally) in order to provide education in the use of e-Infrastructure.

1. Recommendations as to the harmonisation of education in the use of e-Infrastructure.

1. Propose standards for student and teacher identification that would enable access to educational grid facilities and authorization/management of the resources used.

1. Propose standards for sharing training material and t-Infrastructure between institutions.

1. Establish a system for agreeing standards of framing that accredit workers who design, build, operate and support e-Infrastructure so that qualifications are recognised internationally.

6. Future work

7. Contributors

8. IPR statement

The OGF takes no position regarding the validity or scope of any intellectual property or other rights that might be claimed to pertain to the implementation or use of the technology described in this document or the extent to which any license under such rights might or might not be available; neither does it represent that it has made any effort to identify any such rights. Copies of claims of rights made available for publication and any assurances of licenses to be made available, or the result of an attempt made to obtain a general license or permission for the use of such proprietary rights by implementers or users of this specification can be obtained from the OGF Secretariat.

The OGF invites any interested party to bring to its attention any copyrights, patents or patent applications, or other proprietary rights which may cover technology that may be required to practice this recommendation. Please address the information to the OGF Executive Director.

9. Disclaimer

This document and the information contained herein is provided on an “As Is” basis and the OGF disclaims all warranties, express or implied, including but not limited to any warranty that the use of the information herein will not infringe any rights or any implied warranties of merchantability or fitness for a particular purpose.

10. Full copyright notice

Copyright © Open Grid Forum (2006-2008). All Rights Reserved.

This document and translations of it may be copied and furnished to others, and derivative works that comment on or otherwise explain it or assist in its implementation may be prepared, copied, published and distributed, in whole or in part, without restriction of any kind, provided that the above copyright notice and this paragraph are included on all such copies and derivative works. However, this document itself may not be modified in any way, such as by removing the copyright notice or references to the OGF or other organizations, except as needed for the purpose of developing Grid Recommendations in which case the procedures for copyrights defined in the OGF Document process must be followed, or as required to translate it into languages other than English.

The limited permissions granted above are perpetual and will not be revoked by the OGF or its successors or assignees.

11. References

1. National Cyberinfrastructure Council, Cyberinfrastructure Vision for 21st Century Discovery, NSF 07-28, March 2007 (URL http://www.nsf.gov/pubs/2007/nsf0728/index.jsp/
2. See Computing Business News, “Review 2007: IT skills and careers”, Glick, Bryan, Computing 20 December 2007.
3. See Contractor UK News, “Europe faces ‘imminent’ IT skills crisis”, 22 July, 2005 at http://www.contractoruk.com/news/002206.html.
4. See ComputerWorld, “The 8 Hottest Skills for ‘08”, Hoffman, Thomas, 31 December 2007.
5. Thibodeau, Patrick, “The goods for grid”, ComputerWorld, 18 October, 2004 at http://wwwtechworld.com/features/index.cfm?featureID=924&printerfriendly=1.
9. See EC COM (2006) 816 final, Implementing the Renewed Lisbon Strategy of Growth and Jobs, “A Year of Delivery” and EC COM (2005) 118 final, Building the ERA of knowledge for growth, p. 3.
10. See ERA Green Paper, http://ec.europa.eu/research/era/pdf/era_gp_final_en.pdf
16. Contributors to the 2nd ICEAGE Forum concluded that, “the response should be a …systematic development of a professional discipline, with a body of knowledge and professional practices that will lead to reliable, cost-effective and predictable distributed systems projects and operations”, in ICEAGE D1.F2: Second Forum Report, 22/5/07. See also EU Directive 89/48/EEC, which addresses recognition of professional qualifications across EU Member States.
19. Sinnott, Stell and Watt refer to the “fluidity of the technological landscape” so that “grid technology and associated standards are perpetually evolving with new recommendations and software from standards bodies and solution providers”, in Sinnott, R.O., A.J. Stell and J.P. Watt, “Experiences in Teaching Grid Computing to Advanced Level Students”, National e-Science Centre, University of Glasgow.
20. See ICEAGE D1F2
21. Sinnott, Stell and Watt highlight the problem, explaining that “understanding the technical and non-technical aspects associated with security is crucial, not least due to the degree of trust between resource providers and the potentially highly distributed remote end users”, in Sinnott, R.O., A.J. Stell and J.P. Watt, “Advanced Security Infrastructures for Grid Education”, National e-Science Centre, University of Glasgow.
22. Please see JANET Roaming Policy, v. 1.47 – 11 April 2006, http://www.ja.net/services/network-services/roaming/documents/policy.pdf (JRS or JANET Roaming Service, is the UK branch of the eduroam confederation), JANET Roaming home page, http://www.ja.net/services/network-services/roaming/, University of Bristol technical specification compliance case study document on JANET Roaming documentation page (etc.) as well as the university’s acceptable use policies (which includes university-specific regulations for use of computing facilities, code of conduct and JRS usage policy) at http://www.wireless.bris.ac.uk/wordpress/?page_id=20, and European eduroam confederation policy, Version 1.1, GEANT2 JRA5 deliverable “Roaming Policy and Legal Framework Document – Part 2”, http://www.geant2.net/upload/pdf/GN2-06-080v4-Deliverable_DJ5-3_2_Roaming _Policy_and_Legal_Framework-Part2_20060719163405.pdf
24. Oxford University has delved into the issue of IPR in grid computing environments through the IMaGE Project, which particularly focuses on the complexities of sharing medical data. The project has examined eDiaMoND, the UK eScience Digital Mammography National Database, which is being developed through grid technology applications. The IMaGE analysis teases out issues that could be relevant when considering IPR models to suitably frame sharing in EU grid computing and grid education. See D’Agostino et. al., “On the Importance of Intellectual Property Rights for eScience and the Integrated Health Record”, Oxford Projects, IMaGE and http://www.oerc.ox.ac.uk/activities/projects/index.xml?ID=image
25. See Foundation for Information Policy Research, Text of Directive 2001/29/EC at http://www.fipr.org/copyright/eucd.html
26. See the Berne Convention text at http://www.law.cornell.edu/treaties/berne/overview.html and TRIPs page at http://www.wto.org/english/thewto_e/whatis_e/tif_e/agrm7_e.htm
28. ICEAGE: www.iceage-eu.org/library and EGEE: http://egee.lib.ed.ac.uk
31. See https://forge.gridforum.org/sf/wiki/do/viewPage/projects.et-cg/wiki/ExistingCourses
32. See EGI website for current information on NGI development in each Member State:
34. OGF doc
35. See “Towards Professional Grid Certification”, draft doc 14419, ET-CG GridForge, OGF at https://forge.gridforum.org/sf/go/doc14419?nav=1

12. Appendices

Appendix A – Definitions

Education – a long-term process using conceptual models and resulting in development of a culture.

Eduroam – Education Roaming, or eduroam, is an infrastructure that allows staff and students to access wireless networks at cooperating universities across the EU (and elsewhere) using their home institution username and password (so they do not have to set up new accounts at institutions they visit). Access is secure and mobile.

European Grid Initiative (EGI) – a European level infrastructure based on NGIs, currently in its design phase. The EGI will coordinate NGI interaction and integration in order to improve access to resources and research across the EU. Once developed, it would be considered a key element of the ERA.

Grid computing -

“grid in a box” – a local grid infrastructure with easy installation and portability that can be used in universities and other institutions that do not already have developed (or appropriate) grid computing infrastructures.

National Grid Initiatives (NGIs) – national initiatives to connect and expand grid infrastructures in Member States in order to integrate resources for research and allow for coordination, coherence and interoperation for users of grid computing and its applications across disciplines.

Training – a short-term process to develop specific skills in a certain technical area.

Appendix B – Masters and Other Courses in Grid Computing

Appendix C – Grid Education Curricula and Textbook Development Resources

1)Grid Technology Cookbook, SURA http://www.sura.org/cookbook/gtcb

2)GridForce Project Bina Ramamurthy, SUNY at Buffalo http://www.cse.buffalo.edu/faculty/bina/gridforce/first.htm http://www.cse.buffalo.edu/faculty/bina/

3)International Workshop on Collaborative and Learning Applications of Grid Technology and Grid Education, 2005 and 2006. http://gsic.tel.uva.es/clag/clag2006.html

4)ACM Curricula Recommendations: http://www.acm.org/education/curricula.html

SIGCSE, ACM Technical Symposiums on Computer Science Education, 2005-2007 (also upcoming 2008 Symposium, “Diversity through accessibility”, 12-15 March, Portland OR) http://portal.acm.org/browse_dl.cfm?linked=1&part=series&idx=SERIES307&coll=portal&dl=ACM&CFID=21520226&CFTOKEN=81262262

5)IEEE Computer Society Computing Curricula Series http://www.computer.org/portal/site/ieeecs/menuitem.c5efb9b8ade9096b8a9ca0108bcd45f3/index.jsp?&pName=ieeecs_level1&path=ieeecs/education/cc2001&file=index.xml&xsl=generic.xsl&

6)BCS Education and Training Forum http://www.bcs.org/server.php?show=nav.6042