Managing Cyberinfrastructure Strategically

1
Managing Cyberinfrastructure Strategically

• Patrick Dreher
• Director, Advanced Computing Infrastructure and
  Systems
• Renaissance Computing Institute

2
Outline

• Snowmass Meeting
• Campus Cyberinfrastructure (CCI)
  characteristics engines - drivers
• Cyberinfrastructure Summit in Denver
• Follow-on based on the Denver meeting
• EDUCAUSE initiated CI projects
• Community participation - how your institution
  can become more involved in these efforts

3
Campus Cyberinfrastructure Snowmass Workshop

• The meeting was the first workshop for the
  Net_at_EDU Campus Cyberinfrastructure (CCI) working
  group
• Question what are the key components of
  cyberinfrastructure?
• Sponsored by the National Science Foundation
• August 45, 2006 Snowmass Village, Colorado
• http//www.educause.edu/nmm061

4
Engines/drivers of Campus Cyberinfrastructure

• Campus Communities and Constituencies
• Enabler in partnerships with researchers, not as
  an IT service provider
• Computing and Communications
• Opportunities, synergies (grids), economies of
  scale for high performance computing, research
  networks and enhanced support facilities
• Information Management
• Various aspects of data creation, storage,
  handling, retrieval, distribution interpretation,
  security, policies on research data, including
  partnerships and opportunities with libraries and
  repositories
• Virtual communities
• Opportunities associated with scholars
  partnering with IT organizations to create the
  software environments that facilitate discovery
  among distributed communities
• Partnership Strategies
• Development of proposals and relationships that
  enhance the partnership among researchers,
  universities, and funding organizations to
  enhance the nations cyberinfrastructure  (local,
  state, federal, international, private)

5
EDUCAUSEs Grand Challenges Program

• EDUCAUSE is hosting several meetings on topics of
  particular importance to higher education
• This Denver Cyberinfrastructure (CI) Summit was
  part of the EDUCAUSE Grand Challenges Program
• Denver meeting attended by about 50 higher
  education leaders with particular expertise and
  responsibility in the support of IT for research

6
Cyberinfrastructure Summit

• List of attendees http//www.educause.edu/13280

7
Goals for the Denver CIO Summit

• Obtain CIO insights and perspectives for
  generating a set of national CI priorities
• Capture best ideas and strategies for advancing
  CI from this assembled group of IT leaders
• Formulate strategies and tactics for EDUCAUSE and
  its members
• Meeting URL and final report located at
• http//www.educause.edu/cisummit

8
Toward Managing CI Strategically

• Recap of the Summit Meeting
• Highlight/focus on several of the ideas (final
  report contains the full details)
• Summit Format
• Several brief overview talks
• Structured discussions within breakout groups
• Multi-tier role
• CIO level
• Middle layer campus cyberinfrastructure
• Individual principal investigator

9
Overview Presentation Themes

• Cyberinfrastructure
• What is it?
• Why do we care?
• What CI components should be supported?
• Where are the resources?
• Organizationally who actually does the planning
  and secures the funding?
• Next steps?
• CI is a complex mix of components

10
Cyberinfrastructure Players Peter Siegel Talk
Researchers
Faculty
Grad Students
Staff
University Consortia Systems

• Russ Hobby, Internet2

11
Cyberinfrastructure Functions and ResourcesPeter
Siegel Talk
12
Russ Hobby, Internet2
The Network is the Backplane for the Distributed
CI Computer Peter Siegel Talk
Human Support
Training
Help Desk
Education And Outreach
Collab Tools
Publishing
Network
13
Breakout Discussion Groups Five Major Areas Of
Emphasis

• Leadership
• Support for Research
• Priority and Funding
• Short Term Strategies
• Long Term Strategies

14
Common Categories Within Each Breakout
Discussion Group

• Established Practices
• Trends
• Ideas
• Recommendations

15
Leadership
16
Leadership-- Established Practices 1--

• Engagement communicate with executive
  management show CI as a community need, not an
  IT need
• Understand the local environment and culture
• Build relationships with those building all the
  of the rest of the campus physically and
  infrastructure

17
Leadership-- Trends --

• Planning, selling, engagement, and implementation
  tied to the institution's strategic plan
• Integrate CI into university plan
• The CIOs goals must be connected to institutional
  strategy
• CIO must be an equal colleague of those s/he is
  influencing
• CIOs should partner as the architect/rainmaker
  for CI -- not just the plumber or the builder
• Demo successful partnerships
• Active use of benchmarks

18
Leadership-- Ideas --

• Define what CI is on your campus
• Need for collaborative educational effort
• CI is different from past IT initiatives it is
  more externally influenced

19
Leadership-- Recommendations --

• Have a major coordinated campaign, on a national
  level, coordinated by EDUCAUSE
• Some leaders are needed across community
  disciplines to build groups that span and mediate
  needs with actions and resources.
• Influence those who control the money connect
  with the leaders within the leadership groups.

20
Support for research
21
Research--Established Practices --

• The definition of core CI services changes over
  time
• Institutions should implement process to
• Ensure CI meets reasonable expectations
• Adjusts campus investments to shifting
  priorities.
• Researchers, IT leaders, and administration must
  engage in continuous dialogue and review of the
  CI environment, including funding agency plans
  and commitments, to ensure the overall systems
  sustainability.

22
Research-- Trends --

• There is rapid expansion of demand for a robust
  CI on campus to support research
• Key elements needed in planning for research CI
  needs
• The technology itself
• Involvement of central IT
• Workflow, policy and funding proposal development
• Effective IT governance structure.
• Human considerations
• Early involvement, cooperation with research
  faculty
• collaborations and partnership arrangements
• Support and consultation
• System ease-of-use
• Reliability
• Interoperability

23
Research-- Recommendations --

• Establishing life-cycle replacement process
• Run development, test, and production systems
  simultaneously
• Continuous involvement with users
• Balancing leading and bleeding edge
• Keep abreast of international, national,
  regional, and campus developments
• Understand how CI must adapt to accommodate
  cross-disciplinary needs and domain-specific
  research.

24
Priority and Funding
25
Priority and Funding-- Established Practices --

• Funding agencies spawn independent small clusters
  for individual researchers, accentuating problems
  of high cost solutions
• Burdens need to be shared by campus and sponsors
• Focus on engagement Wins for the rainmakers
• ROI arguments used to get buy-in
• Establish common expectations/norms for costs
  (including hidden costs) and benefits

26
Priority and Funding-- Trends --

• CIO beginning to facilitate partnerships across
  all campus sectors
• Leading from the side, advocating
• CIO must nuture contact with researchers or s/he
  is in danger of being just a utility
• Sharing verses owning gt cultural change
• Use central seed funds, partial funding, priming
  the pump
• Functional lead vs technology lead

27
Priority and Funding-- Ideas --

• Develop a commonly funded CI consortium approach
• Demonstrate environmental impact of
  distributed/incoherent approach compared with CI
• Campuses need to organize globally prioritize
  locally
• Practice good stewardship of the whole not just
  local,
• Tap faculty incentive packages to reinforce
  common CI
• Implement CI as a strategic capability for
  institution
• Focus on innovation
• Engage non-vested partners (economic
  development, independent validation)

28
Priority and Funding-- Recommendations --

• Create sustainable funding models
• Request an NAS/NAE study on value of federal
  funding of CI to science and engineering
  research.
• ECAR should do a short-term study on campus CI
  funding models
• CCI Focus Group Report Campus
  Cyberinfrastructure and Data Centers

29
Short Term Strategies
30
Short Term Strategies-- Established Practices --

• Begin to establish social technical trust
• Identify, engage leaders and influencers
• Create incentives for good behavior
• Create common body of materials
• Success failure stories
• Articles
• Research examples
• Events/meetings

31
Short Term Strategies-- Trends --

• Partner CI to faculty initiatives and research
• Begin development of campus, regional, and
  national awareness around CI
• Tangible pilot projects
• Develop technical CI-oriented staffing and
  expertise local, regional, national
• Identify and leverage opportunities between
  traditional IT CI

32
Short Term Strategies-- Ideas --

• Align rewards to practicing good CI to hiring
  tenure practices
• Define CI as an institutional competitive
  advantage
• Real projects (venture capital model)

33
Short Term Strategies-- Recommendations --

• Convene and engage prominent research faculty in
  ways appropriate to your campus culture
• Share outputs to larger CI community
• Search for compelling domain cases for
  highlighting CI/piloting (demo projects).

34
Long Term Strategies
35
Long Term Strategies-- Established Practices --

• Growing importance of data mix public/private,
  library
• Paradigm shift service, collaboration, tools.
• Emphasis on growth of software tools for
  collaboration
• Greater focus on service
• Continued demand for end-user support
• Close alignment with existing institutional
  incentives
• Discipline-specific recognition of unique needs
  and drivers for services and support

36
Long Term Strategies-- Trends --

• The scale of collaboration will increase greatly
  with deepening research collaboration
  relationships
• Between IT and researchers
• Among researchers themselves
• Increased connection to libraries and digital
  asset collections (Information Management pillar)
• CI will be leveraged in areas K-12, learning,
  teaching, relationships
• Manage change through transitions and paradigm
  shifts.
• Existing physical plants need updating
• Understand distinction between domain-specific
  and discipline-specific

37
Long Term Strategies-- Ideas --

• The nature of higher education could change based
  on CI influences.
• The effectiveness of technology will always be
  debated
• Remove non-capital costs from principal
  investigator perspective.
• Virtual Research Parks
• There will always be scarce resources, but
  specifics will shift

38
Long Term Strategies-- Recommendations --

• Develop assessment framework to identify trends,
  usage, and effectiveness of CI
• Assess generality/applicability of research CI to
  broader communities (healthcare, K-12)
• Move in the direction of fostering development

39
EDUCAUSE Role
40
EDUCAUSE Role-- Ideas --

• EDUCAUSE can engage international groups to
  ensure that CI becomes a global focus.
• EDUCAUSE should organize a focused joint meeting
  of CIOs and vice presidents for research.
• EDUCAUSE should engage funding agencies on how to
  effectively fund CI
• Formulate a unifying/mobilizing CI message
• EDUCAUSE has not been viewed as an organization
  with a focus on research computing, but it needs
  to be.

41
EDUCAUSE Role-- Ideas --

• CCI Working Group should organize teams to move
  some of the ideas reported here forward
• Distribute CCI active white paper on CI which can
  be quickly shared by CIOs with campus
  constituencies
• EDUCAUSE should foster a Wiki about CI that can
  be used by interested institutions to post
  inventories of their campuses CI resources and
  services
• CCI Focus Group Report Creating a Five Minute
  Conversation About CyberInfrastructure

42
EDUCAUSE Role-- Ideas --

• EDUCAUSE should facilitate a meeting with
  researchers to discuss CI, perhaps concurrent
  with the annual Supercomputing Conference.
• EDUCAUSE should make extra effort to include
  smaller institutions in the CI discussion.
• EDUCAUSE should establish a speakers bureau on CI
  to provide experts for a variety of forums
  (including the Congress of the U.S.).

43
EDUCAUSE Role-- Ideas --

• EDUCAUSE should develop a list of national
  leaders who are proponents of CI to deliver
  keynote addresses, etc. Draw these individuals
  from presidents, VPs for research, provosts,
  deans, senior faculty, and other administrators.
• EDUCAUSE should create a forum for CIOs and
  senior institutional leaders (such as VPs for
  research) to address CI issues

44
Opportunities for Community Involvement
45
Initial Follow-on Actions from the Denver CI
Summit
46
EDUCAUSE Action Item

• Recommendation
• EDUCAUSE should digest CI work done by others
  (e.g., the NSF report) into a succinct IT action
  plan, using new language to help facilitate the
  discussion among important contributing parties
• Charge
• Write a readable and compelling digest of the
  several long documents now describing the field
  (e.g., NSF Vision document) for the use of a
  broader audience including university presidents,
  campus CIOs and other university executives
• Work with an extended advisory group to develop
  steps and strategies that can be used by
  university executives to move a campus ahead with
  respect to CI.

47
EDUCASUE CI Document Summary

• This group is just beginning its activities
• Outreach to community for document preparation
  feedback
• Advisory committee
• Guy Almes, Texas AM and co-chair CCI
• Alan Blatecky, RENCI
• Jim Bottum, Clemson
• Patrick Dreher, RENCI and co-chair CCI
• Sue Fratkin, consultant
• Dave Lambert, Georgetown
• Clifford Lynch, CNI
• Diana Oblinger, president elect, EDUCAUSE
• Craig Stewart, IU

48
EDUCAUSE Action Item ECAR Study

• Charge --
• Determine present state of CI preparation and
  strategies on U.S. campuses with particular
  attention to funding strategies
• Major focus on research institutions and STEM
  research, but participation also from other
  colleges plus teaching and learning.
• CIOs will probably be asked to involve their
  research VPs in the answers
• Mark Sheehan, ECAR principal investigator playing
  major role in this effort

49
ECAR Study

• Advisory committee is now helping to formulate
  the questions for the survey
• Rosio Alvarez, LBL
• Sally Jackson, UIUC
• Kevin Morooney, PSU
• Jim Pepin, Clemson
• Peter Siegel, UCD
• Joel Smith, CMU
• EDUCAUSE is funding this directly so the results
  can be shared publicly immediately on completion

50
Middle LayerCampus Cyberinfrastructure
51
Campus Cyberinfrastructure(CCI) Working Group

• Campus Cyberinfrastructure (CCI) Working Group
  Open Meeting
• Thursday, October 25, 20071200 p.m. - 500
  p.m.Leonesa I, 1st Floor (Grand Hyatt Hotel)
• http//www.educause.edu/E07/Program/11073?PRODUCT
  _CODEE07/MTG39

52
CCI Meeting Strengthening the middle CI
framework

• Themes for the CCI group meeting
• Where does one identify and support this common
  core middle level of CI
• Interfaces with the national layer and individual
  PI independent of local campus cultures or
  specific disciplines?
• How can we create a palate of different tactics
  that
• Can be adopted to particular campus
• Support our agreed-upon national priorities?
• How can campuses, working through EDUCAUSE,
  implement these strategies?

53
Discussion and Questions