OUCS Annual Report 2006-2007
1. Introduction
Dear Colleague,
Welcome to the annual report of Oxford University Computing Services for 2006-2007. In keeping with previous formats we have attempted to write a report that is informative and approachable to allow our users to get a comprehensive glimpse of OUCS’s performance over the period. We have divided the report into three main sections. First, in an attempt to synthesize issues which have arisen, we have produced a set of short essays on current ‘perspectives’; namely topics which have come to the fore over the last year. We then present a set of service reports which show the increasing levels of demand on OUCS. This year, where possible, we have also attempted to show the level of use of OUCS’s services by the Divisions. Finally, we include a series of reports from the Research Projects OUCS has led or been part of over the year. This aspect of OUCS is fundamental to our taking services forward for the University, and we attract over £1m per annum from external funding agencies.
As well as undertaking major initiatives to develop our services in line with the demands of our users (e.g. the upgrade to Herald), 2006-7 witnessed some important strategic changes for OUCS. First, the University’s ICT Strategy was finalised and the first steps have been taken to introduce a new governance structure. The University’s ICT Committee was disbanded to be replaced by the PRAC ICTC sub-committee (to meet for the first time in Michaelmas Term 2007); and a new professional group for the IT Support Staff across the University, whose invaluable work we all rely on, was set up. More importantly, the Office for the Director of IT has been established, looking at strategic IT initiatives across the University. Throughout this period OUCS has continued to develop but the lengthy period of time taken to ratify the ICT strategy has not been without consequences, with several senior managers being asked to ‘act-up’ during the prolonged period.
Equally important has been the setting up of the Services Funding Working Group (SFWG). This was a central committee with divisional representation charged with looking at the funding of central services – including OUCS. We liaised with this committee throughout 2006-2007 and have now started to implement its recommendations. For the first time the budget of OUCS will be set to meet the increasing demands of the divisions, and anomalies in our funding (which over the years has led to a mounting deficit) are being tackled. Long-term planning and funding of key services will be addressed, and future capital projects properly resourced. The SFWG also revealed that although OUCS is solely funded by the Infrastructure charge levied on the Divisions (in effect to provide services to c. 26,000 students and staff) we are actually supporting over 35,000 users across the collegiate University.
OUCS has also made major strides to meet its users and to get feedback from them on our services. We have a series of OUCS Reps who are linked to colleges or departments. These provide a direct route for feedback to us primarily from the Academic staff in the departments, faculties, and colleges to supplement our existing communication network with IT Support Staff. OUCS staff also sit on all the major Divisional IT Committees. In keeping with the times we are now using Web 2.0 technologies to facilitate communication, including an in-house developed news feed service (OxITEMS which now facilitates Podcasts), Facebook groups, and blogs and wikis.
Finally, it should be noted that this Annual Report forms only part of the information OUCS provides. Here we attempt to indicate how we have performed over the year; but we also produce a five-year strategy looking forward (revised each January). Furthermore, this year we have produced for the first time a complete set of Service Level Descriptions providing information on all the services we provide. All of these are available from the OUCS web site (
I hope you enjoy reading this report and find it informative,
Dr Stuart D Lee
(Acting) Director, OUCS
2. Perspectives
In this section we attempt to present short essays on emerging trends and technologies that OUCS has been involved in over the year. We have built up links with departments and colleges and constantly look to investigate technologies in answer to our users' needs. The following sections highlight some of the areas we have been focusing on.
2.1. Reducing the IT Carbon Footprint
The University of Oxford has set up an Environment Panel to advise on the environmental implications of University policies and activities and to encourage the proper consideration of environmental factors in the University’s decision-making processes. One of the factors the Panel advises on is the reduction of energy consumption within the University. This obviously has an impact on the way we currently use ICT and will influence future plans, but more information is needed about how and why changes can be made.
2.1.1. Low Carbon ICT Project
Since October 2007, OUCS has been working on a JISC-funded project entitled Low Carbon ICT. The purpose of this is to provide practical examples of how an institution can implement both the policies and the technologies to conserve energy used by desktop and server hardware with the minimum of disruption to users and system administrators. This will provide a important case study and forum for other organisations looking to adopt similar strategic goals. The project will feed directly back into the University's environmental panel to further enhance University environmental policies.
By using new technology and changing computer management regimes, it should be possible to significantly reduce the University of Oxford’s ICT carbon footprint. Potentially Oxford’s annual power use could be reduced by a minimum of 2GWh, saving c. £200k, and preventing the production of over 1,100 tonnes of carbon dioxide each year.
It is important to take a holistic approach to develop an environmentally sustainable ICT strategy because savings in one area can have a negative effect in another. For example, poorly designed energy saving measures could compromise security through interrupting essential software updates. It is also important that the core strategic goals of the University are not unduly impacted, namely to conduct research and teaching. It is essential to bring together technical, research, communications, and training skills from across the University to ensure this necessary broad perspective is adopted.
Two strategies for saving energy and reducing the Universities carbon footprint are discussed further here - Wake on Lan (WoL) and Virtualisation.
2.1.2. Wake on Lan
One of the main aims of the Low Carbon project is the widespread introduction and uptake of Wake on Lan (WoL) to Oxford.
For end users, WoL technologies allow people to switch off their desktop machines when away from them for lengthy periods, whether overnight, at weekends, on holiday, at a conference, or even when in a meeting or laboratory all afternoon; but restart them locally or remotely if needed. With normal switching off, someone has to press a button on the machine itself to restart.
For system administrators, WoL facilities will enable them to switch clusters of machines on and off depending on the department's energy saving policy and requirements to update system software on desktop machines. For example, an administrator may switch off a teaching cluster of machines over the weekend then switch them on to install a security patch.
In short, WoL facilities allow the energy saving advantages of switching machines off with the ease of access and update of machines being continuously on.
Recent data from one University department shows that their energy use shows only moderate daytime peaks. This is indicative of a building with its computers on 24 hours a day. The figure below shows the current demand pattern compared with the lower demand pattern that might be expected were the computers to be switched off overnight. The overall electrical savings could be up to 20%. With machines being switched off at weekends, and other planned absences, the possible savings are even greater.
Figure 1. Dyson Perrins Electricity Demand – summer and weekdays
2.1.3. Virtualisation Technology
The NSMS section of OUCS has recently introduced a virtualisation service that enables system administrators to consolidate their existing servers onto centrally managed machines. By exploiting virtualisation technology, it is possible to support 10-15 servers (Windows, Linux, or Netware) on two physical servers. This can save on capital costs for hardware and also provide enhanced management options that can significantly increase the resilience to failure of the IT set up.
As far as the Low Carbon Project is concerned, the benefits of this approach come through minimizing procurement, which reduces the environmental impact caused by the manufacturing and transport of new server machines. The project will monitor the procurement practice of units throughout the university to measure the corresponding reductions on the environmental pressure due to less manufacturing and transport. Fewer physical servers will also mean lower demand for specialist air-conditioned rooms, which consume large amounts of energy.
2.1.4. Evaluating Project Performance
Overall the success of the Low Carbon ICT project will be measured by the amount of energy the University saves as a result of rolling out WoL management facilities and the increased use of the current virtualisation service.
2.2. Third-party IT services
In the dim and distant past (e.g. thirty years ago) OUCS was the sole provider of computing services to the University. In the recent past (e.g. fifteen years ago), other parts of the University began to provide their own computing services of various kinds. Now we are moving into a third age in which computing services of various kinds are available to the University from many different providers. A department, college, or other unit, might decide to use authentication services provided by OUCS, but to run its own mail server, and to host its websites with an external service provider unaffiliated to the University. Increasingly, major IT services such as web hosting, distributed filestore, digital archiving, groupware solutions, and many niche software applications are being provided as commodity services, using a variety of business models. A unit might, for example, decide to manage its projects with a tool such as Basecamp, to share documents amongst its members using Google Docs, to archive digital images using Flickr, or even to recommend Hotmail to its students. This richness and diversity of choice makes this a good time to be a user of such services, but also provide challenges for those responsible for the University's strategic direction and long term interests.
As a small contribution to that strategic thinking, OUCS has prepared a check list of considerations which we recommend should be taken into account when evaluating ways of supplying any IT-related service, in particular where that service might be “contracted out” to a supplier outside the University.
The check list is reproduced below and can also be accessed online at:
Availability and reliability
What guarantees are provided about support or level or service? Are they adequate for the intended use? For example, if the service is unavailable for an extended period of time, how seriously would this affect the unit's activities?
Continuity of service
Is there any agreement concerning continuity of the service? How serious is the risk that the service might change its policies, or prices, or go out of business? For example, does the unit care if a free service becomes a paying one, or one subsidized by advertising? Does the unit have an alternative or exit strategy in such an eventuality?
Support issues
What level of support activity will be needed? Is the service widely used by comparable units outside Oxford? What is its public perception? Is there a strong community of existing users who can provide peer support, or will the unit need to seek or provide specialist training? Does the service provide (e.g.) a hotline for academic or technical support issues beyond the run of the mill? Is it likely that existing IT support networks (for example OUCS Help desk) would be able to provide first line support?
Migration issues
If the proposed service overlaps wholly or in part with a service already used by the unit, what will the support costs be in moving existing users? Is migration without loss of information simple, technically feasible, or impossible? Does the proposed new service have all the functionality of the existing one (for example, does it provide the same or enhanced levels of security, backup, etc.)?
Domino effects
Will introduction of the new service affect other existing services, for example by reducing or increasing their importance or requiring changes in them? Is there a risk that the service would increase for example network traffic or spam beyond currently acceptable thresholds?
Duplication effects
Is the service (or something analogous) already being used by some other unit? If so, is there scope for co-operation for example in licensing or in pooling of support activities? Alternatively, is there a risk of confusion or lack of data integrity if the same service is provided under different brandings within the University?
Strategic and legal considerations
How well does the new service conform to established University strategic priorities or practice? For example, can it take advantage of current university-wide authentication and identification systems? Is it equally usable in all hardware and software environments of importance to the unit? Does the new service open the unit to possible additional risk with respect to its legal obligations, such as privacy legislation, or contracts (JANET regulations)? Is the service provided under terms specified by an enforceable contract between the service provider and the University?
Rights issues
Are the terms and conditions appropriate for the intended use? For example, if the service will store or manage material in which the unit has rights, do the terms and conditions adequately protect those rights?
Privacy and confidentiality
Are the terms and conditions adequate for the intended use? For example, who may access the unit's data and in what circumstances? Is usage of the service auditable by the unit (for example, to track any alleged abuse)?
Cost implications
What are the cost-benefit implications of using the new service? What is the total cost (or saving), taking into account all the above considerations, of using this service as opposed to expanding (or continuing with) an existing internal service or doing without? How will that cost be met?
We believe this checklist provides a good basis for the evaluation of 3rd party IT solutions, and have applied it ourselves when considering an externally-provided groupware solution for use within OUCS.
2.3. And the Survey Said…
We often hear about the future being mobile, secure, ‘anywhere anytime’ computing but do we have really know what technology our university students are currently using in their day-to-day lives? Over the last four years OUCS has conducted annual surveys of new students at the Freshers’ Fair in order to ascertain their IT ownership and ICT skills. This data gives information on trends among the incoming student population to help inform future service provision.
As a typical example, the 2007 survey details were based on 621 students. The age range was between 19 and 24.
Figure 2. Student-owned Devices
Laptop ownership far outweighs ownership of desktop computers and almost all students have a mobile phone. Memory sticks, MP3 players, and digital cameras have seen steady increases over the four years of the survey.
Of these computers, PC/Windows machines are still the most widely used platform but are at their lowest level since the survey started in 2004. Mac and Linux machines are increasing and account for 15% (from 8% in 2004) and 3% respectively.
In terms of ICT training, the most popular requests were for training in databases, graphics programs, and programming. Graphics programs are the only category to show a consistent increase, possibly due to the rise in digital camera ownership.
Figure 3. Recreational Use of the Internet
When asked what they use the internet for, most fresher students said e-mail communication (more than 90%), followed by those who use it for general surfing websites and chat (both over 70%). The most dramatic increases are in the use of chat and sharing photos.
A question on the use of social networking sites was added this year, and the results show the Facebook service has reached almost universal coverage (96% usage) and obviously is now as popular out of University social circles as within.
Figure 4. Online Communication Methods
Email and Messengers are the two preferred communication channels between students. A significant increase has occurred in the use of internet telephony services such as Skype conferencing software and the (possibly related) ownership of webcams.
2.3.1. Conclusions
With their laptops, mobile phones and MP3 players, coupled with easy access to wireless facilities, students are ready for ‘anywhere anytime’ computing. OUCS needs to maintain and enhance its facilities, training, and advice relating to podcasts, e-learning, and web pages suitable for accessing by small screen devices. More wireless access points and easy authentication to University services are required.