3. OUCS's rôle in the University
The Mission statement (http://www.oucs.ox.ac.uk/about/mission.xml) defines OUCS's rôle within the University. This section outlines some of the important developments which are underway or planned as part of the ongoing development of the department.
The demands for higher bandwidth will continue to grow — the rule of doubling every 18 months remains valid, although there will be surges as major applications come online. It is anticipated that Grid computing will place large traffic loads on the network, as will the (long predicted) multi-media applications. Networked storage is still in its infancy but is likely to become a major player within the five years.
Plans (and SRIF2 funding) are in place for the next round of enhancements to the backbone network, in 2004. Meanwhile, industry is already developing the next generation of equipment, and it can be expected to become a commercial reality within the next five years.
Raw bandwidth alone is only part of the requirement, and has implications for other services that are run. For example, ‘Firewall’ speeds have managed to keep up with the current bandwidth requirements, but there is a considerable risk that this may not continue to be the case as network speeds increase, and present security techniques will be unable to keep up.
With IP now being widely seen as the transport of choice for many different services, each with its own characteristics, the network will evolve to handle these diverse requirements. Use of non-IP protocols will continue to decline as suppliers focus their resources on IP developments at the expense of these other protocols. This shift of emphasis is already visible, and the University must follow.
Email will become the standard and accepted route for formal, as well as casual, communications within the University and it will become ever more important that all members of the University have easy access to their mailboxes at all times. It will become part of a ‘communications environment’ that will be demanded in the future; in particular, its integration with a future IP telephony infrastructure will be seen as important (ref. 184.108.40.206). This will also require consideration of regulations concerning the responsibility of members to ensure that they have regular access to their email, the preservation of email, etc. OUCS frequently faces the problems that affect the users of externally-hosted mailboxes, and these are unlikely to cease, so the provision of a central mail server/storage resource will continue to be essential. Standards will be required for email servers and clients to ensure adequate handling of attachments, but this should present little difficulty. An increasing reliance on email will require further consideration of security implications, in particular with respect to the forging of email headers, and it is likely that there will be increased use of public key cryptography and digital certificates in signing/encrypting emails.
IPv6 is the ‘next generation’ protocol, designed by the IETF (http://www.ietf.org/) to replace the current version Internet Protocol, IP Version 4, which is used on the University networks. IPv6 adds many improvements to IPv4 in areas such as routing and network auto-configurations, and although it is still under development, the major network suppliers are starting to introduce IPV6 capabilities into their equipment. OUCS will expect to introduce IPV6 within the next five years. This will present major technical challenges and OUCS will oversee and coordinate a cooperative programme of implementation throughout the University. It is expected that IPV4 and IPV6 will have to coexist for many years, and OUCS will be required to provide and support gateway equipment. Experimental work on this is just beginning.
There will be a large demand for location-independent methods of connection, which probably will lead to a standardised way of wireless networking across the University. In the long term, wireless networking may be expected to become the main method of connectivity (for voice as well as data), with fixed connections only being needed for high-bandwidth applications. There will also be increasing demands for ‘home-working’, and interconnection services, such as Virtual Private Network, will continue to be vital. Security considerations will be of paramount importance. Wireless access points, for example, should be considered as outside the trusted network, and may mean that trusted users will have to ‘tunnel’ to the trusted network (unless a reliable authentication system can be developed).
The proliferation of access to computers is now becoming a key issue. It is important to consider how this will affect student learning and teaching practices. Hand-in-hand with this is a possible IBM ‘Thinkpad’ project where laptop/tablet PCs will be made available to students at competitive prices and will lead to more uniform hardware and software provision, and concomitant support benefits. There may be a widespread take-up of PDAs and tablet PCs, and it will be necessary to assess how best these can be harnessed to support student learning and teaching.
Discussions are being held between OUCS, the Telecommunications section and Security Services over possibilities for connectivity and future requirements. It is important that the groundwork is laid now so that the University has a coherent framework on which to develop its telephony services. Its integration with the rest of the demands on the network must be fully considered if the ‘communications environment’ is to be successful, and the total provision of cables and fibres throughout the University must be reviewed.
A Network Management Working Party has been set up recently to address this area, which will give the University, for the first time, a body in which these issues can be brought together and future planning coordinated.
The future of Janet as a general-purpose provider will be increasingly questioned, as ISPs become accustomed to supplying reliable and high-speed services to meet business demands, and widespread provision leads to realistic pricing. Provision from more than one supplier offers the potential for protection against failure, although this is not straightforward and has major implications for the way the University achieves its internet access. It would, however, enable the University to support ‘commercial’ activities and collaborations not allowed over Janet at present. A separate but overlapping question is that of the College connections, which are often heavily entertainment-orientated, and are imposing an increasing burden of policing and legal responsibility on the University. Other universities already adopt the policy of using a commercial supplier for student residences; this is of course complicated at Oxford by the dual work/residence nature of the Colleges.
A generation of SuperJanet, to support very high-speed research applications without being clogged by general traffic, may come about, as in the US. This has serious political implications: if it is not funded for all universities, it immediately creates a virtually unbreachable two-tier system, and would almost certainly result in increased costs for the University.
The University network, and the machines attached to it, will continue to be a prime target for hackers/crackers, and new developments in data sharing will bring new opportunities for security breaks. The rôle of a central body to monitor the University systems, to respond to incidents, and to be a source of advice and education for systems managers, will continue to be vital.
As the internet races forward, the law runs to keep up, and the providers of University services must always be aware of the legal risks and liabilities that surround them. Much effort is devoted to handling threats of action over breach of copyright (defamation is another common concern). Under current understanding of the law, the University is not so much at risk from breaches of civil law by its members, but more for not responding quickly to complaints about such breaches. This is another rôle that needs to be managed centrally, while OUCS will contribute fully.
The issues concerning data preservation and curation will have ever-increasing importance. OUCS will continue to play a major rôle in all these areas and respond to changing demands from departments. The HFS service is currently being upgraded to provide increased performance, resilience and greater storage capacity.
The HFS service may be considered in two main functional areas: backup and archive. The distinguishing feature is that the former records the changes in data that is being actively used on the system being backed-up, so that it can be restored if that system is lost or damaged, while the latter preserves data in the longer term, when it is no longer being developed.
Provision for the day-to-day backup of working is considered to be vital, and the University must always strive to minimise the danger of neglect of this need. It is currently accepted that a central back-up service, free at the point of use, is the only way to encourage adequate safeguards being taken by individuals. Larger departments may have their own policy, often supported by the central service.
Preservation of data is essential to the University (and to any major organisation). The data that has to be considered covers a wide range, and includes data generated by research projects, archives of University business, and digitised representations of assets such as library and museum holdings. All these areas are expected to grow rapidly, as the data generated by research projects grows ever-larger, as all new printed material becomes available digitally, and digitisation of historical material becomes cheaper.
One of the first steps of any archive provision must be to establish a method of determining what is worth preserving (for backup data, it can be assumed that as someone is working on it, it is of importance). The fundamental point is that preserving data carries a cost, and that cost is the responsibility of the owners of the data, whether the University as a whole, one of its departments or the funders of particular research.
At the moment, most emphasis is given to storing material, but the difficult part is making that material available on demand. The thrust for better meta-data handling will continue, but it is not clear that general solutions will emerge. The JISC Digital Curation Project is an example of current work that may reap major benefits in the future (and Oxford must decide whether to bid to host the new Centre). Demand, led by the Grid and other huge data collections, will lead to viable solutions to the authentication and security considerations.
The problems of data handling are looming large in many areas of business, administration and research, and significant advances in methodology are expected- for example in methods of determining changes in data, and so minimising the amount that has to be backed-up (data coalescence). These are areas where technology is advancing rapidly, for example with network-attached filestore, and SAN technology, the goal of which is to enable heterogeneous data sharing and cross-platform connectivity. Work in this area will make it easier to handle data in a structured manner, with departments and OUCS able to share responsibility.
OUCS would like to extend its rôle in the area of data management, and to provide a wider service to the University. Consideration should be given to a model for a University-wide filestore, supported and backed-up centrally. This model has applications in a Grid environment, where it could be used to enable and facilitate transparent access. A development on these lines would provide new and easier ways of working throughout the University, could be a driver for cross-disciplinary projects, and would be of major assistance in facilitating common access to data, both administrative and technical.
OUCS operates a range of services on behalf of the University, the most prominent of which is the provision of web server facilities to support many departmental, college and individual web pages. There is a tension in web development and presentation between the exercise of departmental autonomy (especially where departments require that their web pages operate as part of a national or international structure within their discipline), and the need to present a consistent interface to the University, conforming to legislative requirements. It is expected that the need for centrally provided web-services will continue to increase. How the legislative issues should be handled is currently the subject of much debate, and although it is likely that this will become the responsibility of central administration, there will be requirements on OUCS to promote, teach and support standard solutions.
IT is part of the fabric of the University, and the importance of e-learning has risen (as witnessed in the forthcoming University Learning and Teaching Strategy). Oxford has a growing number of initiatives related to ICT in teaching and learning and has produced recent reports and recommendations which are currently being enacted. Some examples are:-
- The OxTALENT survey into IT in Teaching and Learning
- The EPSC/ICTC Working Party on IT in Teaching and Learning
- Divisional strategies — notably from the Medical Division (published and now being enacted), and also from the Humanities Division
These not only reflect the rising interest in e-learning at the individual academic level but also the recognition that it is key for Oxford to engage in this at a higher level. It is anticipated that the activities of the LTG will continue to become more and more central to the support of teaching and learning across the University.
- developing computer-based teaching and research packages in collaboration with Oxford staff and external bodies as appropriate;
- researching the latest developments in the use of C&IT in traditional university teaching and student learning;
- in collaboration with IAUL, providing training of staff and students in IT literacy skills, plus courses on how to effectively use C&IT in teaching and research;
- disseminating information about the use of C&IT via the website publications, workshops and conferences, and acting as OUCS's formal liaison and point of contact;
- providing user facilities for staff and students to work on projects;
- providing access to specialized learning and teaching applications, and multimedia equipment.
The context within which the LTG operates extends beyond the University and is affected by external factors. The LTG must be aware of external pressures and be responsive to them, primarily to guide University policy to meet the new demands. An example of this is the growing emphasis on e-learning at the DfES which has impacted HEFCE policy and funding resources and strategic directions undertaken by JISC and the Research Councils. RAEs, and to a greater extent TQAs, have all had to be taken into account, as have such recent initiatives as Personal Development Plans and the TQEC paper. Moreover external demands on the University, such as the forthcoming Institutional Audit, have had a direct impact on the priority attached to e-learning.
More and more emphasis (both locally and nationally) is being placed on the promotion of ICT in traditional teaching and learning. In particular measuring ‘effectiveness’ appears to be the key topic emerging from the Department for Education and Skills (DfES), and it is recognised that it is extremely challenging to measure this accurately.
There is an increasing recognition that educational best practice is often overlooked in the implementation and development of learning technologies. It is important, therefore, that the LTG covers this area effectively, in collaboration with IAUL and the Department of Educational Studies.
IT skills are now perceived as a basic literacy on par with reading, writing, and arithmetic. Any IT training programme, therefore, must be in a position to cover a comprehensive range of skills to the level required by the subject and future employers, and must be able to adapt quickly and easily to emerging technologies. Courses will need to change to reflect the working practices of academics, for example shorter more targeted sessions are emerging as very popular. The LTG Advisory Group provides input from divisions specifying the directions in which the courses should develop.
As the number of IT-based services proliferates, there is increasing need for a single and reliable method for authenticating users. There is also a clear requirement for service-specific authorisation data to be maintained in a compatible and standard way. OUCS has been for many years at the forefront of University activities in this area.
OUCS plans to develop a single authentication mechanism for all of its own services, which can be used by other service providers across the University, should they wish to do so. This will be based on the best currently available technology, and will build experience in developing local web-based services and also on the outcomes of the nationally funded DCOCE [Note: The DCOCE Project which formally commenced on 1 Jan 2003 and which will run for two years is expected to play a significant part in the development of interoperable access management services for authentication and authorisation within the University and beyond. It begins at a time when access management is emerging as a crucial development area both locally and nationally. At the national level, the JISC-funded national authentication service, currently provided by Athens, will undergo significant changes over the next couple of years to ensure access management within the UK is based on a combination of open standards and models of devolved authentication. RTS staff have already made major contributions to these developments via their participation in the JISC's Subject Portals Project and will continue to do so via their inputs to the AHDS and the national e-Scienceprojects. Locally, ISG staff in Registration and Databases work closely with all agencies within the University developing single sign-on and authentication facilities, to which they contribute unique experience of the complications of current university wide information sources. ISG staff in Help and Information Services will be early adopters of new authentication systems being developed within the OUCS Unix team and are well placed to evaluate their usability.] project.
‘Single-sign-on’ has obvious convenience benefits to the users of IT services. It also has many implications, both positive in that it facilitates customization and personalisation of services, and negative in that it requires greater vigilance and more complex technologies to avoid compromising the security and privacy of users. OUCS expects to continue to play a major rôle in addressing these issues, both by contributing to the wider policy debate and by providing specific technical solutions as need arises. A desirable goal is to eliminate the use of ‘clear text’ passwords for access to central services, and to replace with a secure single-sign-on solution.
Over the years, OUCS has built up considerable expertise in methods of aggregating authorisation information of different kinds from various sources across the university. As the University develops better integrated sources of such information (most obviously in the ISIDORE project), the need for this activity should decline. However, there will still remain a need to make such information available to the many IT-based services proliferating across the University in an open and standard way. OUCS's intention is, therefore, to focus on the development of tools to facilitate such directory services, which will complement the authentication mechanism mentioned above.
With an effective authentication service, it will be possible to provide access to web services that require user-level authentication. This clearly leads to the concept of a University portal as the common front end to such services. As an interim measure, OUCS is developing a prototype portal into which its own web based services can be integrated. It will also open up opportunities for OUCS to develop further web services, building on the infrastructure being developed. This will include the PC Maintenance service and printing services.
A portal offers personalized and customisable access to distributed information sources, at local, institutional and national levels. The long-term goal will be to maintain and further develop the expertise gained, with a view to making a significant contribution to the evolution of Oxford's institutional portal.
The VLE Pilot Project is now underway with a view to completion in August 2004. If it is successful, and the take-up of a centrally supported VLE becomes widespread, then this will become a major OUCS service/activity in the future, covering not only maintenance and support of the system, but also its development to reflect the Oxford context, its integration with other systems, and associated training.
As noted earlier, the emergence of the VLE, ISIDORE, OSIRIS, and the existing myriad of systems developed by departments and colleges to cope with various administrative procedures, need to be interoperable. In essence then, Oxford is recognising the need for what is being termed nationally as an MLE (Managed Learning Environment). These are becoming ubiquitous, and reflect the clear and present need to provide systems that are cohesive to administrators, academics, students, and external auditors. OUCS has identified this as an area for future investigation.