Skip to Content
UK e-Science

All Hands Meeting 2009

« Past, Present and Future »

Theme 8: User Engagement

User Engagment OR Can we make it easy for users to access the right remote compute resources?

Led by Jeremy Yates (j.a.yates@ucl.ac.uk), Caroline Gardiner (Caroline.gardiner@bristol.ac.uk) and Bruce Beckles (mbb10@cam.ac.uk)

If we want more researchers to do more with their ideas and data, we MUST make e-research services such as compute clusters, storage and visualisation much more accessible than hitherto. In particular we need to address the needs of users from new areas such as Biomedicine, the Arts and Humanities and the Built Environment.

It should be noted that the chief aim of most campus based services is to to help researchers make the move from doing all their research on a desktop PC to making effective use of remote clusters. This is a very difficult step for most researchers.

The barriers to the take up of e-research services by new users are well known, but it is always useful to summarise these in terms of what the users themselves say:

1. "I can do everything on my PC"
2. "It will take me too long to get my job up and running on a cluster by myself"
3. "I use a windows PC and it's all in linux"
4. "I can't program or use a script"
5. "I need storage, not cores"
6. "I need somewhere to render my data"
7. "What I do is not HPC"
8. "It's too complicated to use"
9. "It never works"


The University Sector is building up an increasing corpus of knowledge and best practice, represented by the new HPCSIG Focus Group - Research Facilitation - which is led by Caroline Gardiner (Bristol). It's clear that a local approach that employs people whose role it is to facilitate research programmes that use our e-research services is now bearing fruit.

The broad aim of this session is to state our collected best practice and discuss a strategy that combines the best practice and human resource of localism (best represented by the HPCSIG) with a potential resource and enabling role for the NGS.

Specifically we will discuss:
1. Local best practice and use cases:

  • The role of shoe leather
  • The efficacy of one-to-one sessions
  • Really Facilitating Research: Pro-activism at the campus level
  • What makes a good Research Facilitator?
  • How do we measure success - possible metrics

2. How can the NGS and OMII help - "Facilitating the Facilitators"

  • Building and maintaining code repositories
  • Providing Gui job submission templates for workflow
  • Enabling remote job submission and resource capture from my desktop
  • Keeping the middleware robust
  • Installing the middleware
  • Setting up intra- and inter-campus Grids
  • Making services "windowsy" - the road to real growth?

4. Training

  • What can we share?
  • Where do we have it? Local is very important
  • How do we advertise it?
  • How do we pay for it?
  • What are its benefits?

5. Organising ourselves - the Local vs Centre issue - or is it really a problem?


6. Storage, Curation and Viz services

  • Local services and central enablement?
  • Are these the respective roles of campuses and the NGS?
  • Can we make the use these new kinds of service easy?

This discussion will produce a report for the AHM. It is envisaged this report will be used by HPCSIG members and the NGS to write a common business case for a Local and National User Engagement Strategy.