Author: JISC Communications and Marketing Team
Publication date: 8 October 2007
Publication Type(s): Briefing papers
JISC theme(s): Network
Committees: JISC Support of Research committee
Services: JANET, JANET CSIRT
A study of network provision to internationally rated UK research
departments shows that researchers may be suffering as a result of
inadequate or problematic network provision. This briefing paper
highlights these issues and their consequences and suggests steps to be
taken. It also looks at Grid technology take-up and provides an overview
of how researchers’ work patterns might be improved if unlimited network
bandwidth and adequate resources were available.
The key issues
Network infrastructure
Issue
There is an alarming amount of legacy networking in institutions.
As a result of under-resourcing, around one third of institutions still use
the oldest and most basic kind of network technology that allows
connections at less than a tenth of the speed possible. As about 80% of
institutions have many of their research users connected at these slow
speeds, there is appreciable scope for upgrading these links.
Consequences
Maintaining an association between network upgrades and building
refurbishments in order to try to save money has had a long term negative
impact on research workers and hence on research competitiveness. The
consequences extend from poor responsiveness when accessing web pages to
influencing users to operate poor data backup procedures in order to
minimise network use.
Recommendation
The budgets needed to upgrade departmental networks to current
speed norms are not large so this issue should be examined by institutional
senior management wherever legacy networking is present. Institutions in
which researchers are served by old-style, slow cables and by shared
network access should ensure that they possess and implement plans to
replace this with current technology.
Data backup and storage
Issue
The vast majority of researchers use personal external media (such as CDs
and DVDs) to back up and store data because of inadequate quotas offered on
institutional resources, or the cost of central storage and backup.
Information and communications technology (ICT) service departments appear
to overestimate considerably the role played by central servers to back up
research data and to underestimate the level of use of personal media for
this purpose by researchers.
Consequences
Until the overall situation improves a lot of valuable research
data must be considered to be at unnecessarily high risk.
Recommendation
Institutions must recognise the value of the data being produced by their
researchers and should ensure that institutional policies on the protection
and preservation of informational assets establish systematic approaches to
backing up research data. The potential for shared services for research
data archiving is a useful development currently being explored by some
organisations, while establishing data centre services is another area for
consideration. Currently, under full economic costing, unless there is an
exceptional demand on resources, it is assumed that the storage
requirements for research projects are met out of the overheads for the
project. Institutions may need to re-examine this assumption and allocate
further resources accordingly.
Firewalls
Issue
Centrally managed firewalls are the norm in institutions but can cause
difficulties of varying seriousness, including constraints upon laptop use
by visitors.
Recommendation
There is already comprehensive guidance on network security design produced
by JISC and JANET CSIRT (JANET's Computer Security and Incident
Response Team). Institutional firewall policies should be determined by an
appropriate assessment of risk, bearing in mind the growing body of roaming
workers. Being able to successfully host a flow of visitors is an important
capability of a modern research department, and a key part of this is an
ability to rapidly make visitors productive by allowing their laptops to
access the resources they need.
Management/planning
Issue
Responsibility for network provision differs between institutions, and
perceptions of who is responsible may also differ within institutions, from
department to department. Arts and humanities departments and ICT service
departments tend to view managing and funding the network as a central
responsibility, whereas science and engineering departments tend to believe
responsibility is concentrated in the department.
Consequences
These differences suggest a lack of communication or a lack of clarity
about what arrangements are in place and highlight the difficulty of
implementing an effective and coherent planning process.
Recommendation
The network should be considered to be part of the infrastructure of an
institution. It follows that central planning is required to ensure that
network provision needs are adequately planned for. These planning needs
include scheduling a replacement cycle for networking kit and ensuring that
the network is considered at the design stage of any new builds and
refurbishments.
Clustered computing and the Grid
Awareness of the existence of the Grid is largely associated with science
and engineering departments. Typically these Grid-aware departments possess
their own support staff, their own servers with developed storage and
backup regimes, and a wider variety of technical needs.
Among these departments there is wide use of locally clustered computing
and, while around a third of the Grid-aware departments expect to make
appreciable use of the Grid within two to four years, a substantial number
of potential Grid users have consciously decided to hold back from
involvement because of the learning curve involved and the availability of
their own clusters.
These departments can be expected to continue to develop their own high
powered computing (HPC) clusters, although they suffer from inadequate
resourcing. The growth of HPC will inevitably tend to lessen the
comparative attractiveness of Grid resources. The latter will therefore
need to grow at least at the same rate if they are to remain as an
attractive option in the short or medium term.
Other issues with the Grid include:
-
The Grid community can appear rather closed to those outside it
-
As the Grid is basically a concept, it remains imprecise to some
-
Unless a balanced cross-section of information is available about Grid
computing, some potential adopters will delay putting time into it, while
others may experience disappointment if it turns out to interrupt
research output
Recommendation
JISC recognises that more education is required to assess Grid technology
and plan for its adoption and plans to address this issue.
‘Blue sky’ thinking
Access to unlimited network bandwidth would change researchers’ working
patterns, with improvements including:
-
High resolution viewing of artefacts held remotely
-
Improved working from home
-
Enhanced video conferencing services
-
Access to specialist facilities that are linked in real-time over the
network
-
Wireless access to the network
-
Allowing specialists to be part of several projects in several
institutions simultaneously
-
Improved teleworking by researchers in non-UK universities
Implication
There are support needs to be met if the capabilities of SuperJANET to
deliver improved bandwidth are to be realised within the contexts of these
new working patterns. These support needs will require adequate resourcing
if the benefits of SuperJANET are to be fully exploited.
Further information and resources
This briefing paper draws on a JISC report: Review of
Network Provision for Research Needs by Jon Duke and Andy Jordan. The
study was conducted in the context of the imminent commissioning of
SuperJANET 5 and considered the quality and performance of the network
infrastructure between the SuperJANET 5 Points of Presence and 284 5*
rated departments.