Transcript Cyberinfrastructure Begins at Home Craig A. Stewart ([email protected]) Executive Director, Pervasive Technology Institute Associate Dean, Research Technologies Associate Director, CREST (Center for Research in.

Cyberinfrastructure Begins at Home
Craig A. Stewart ([email protected])
Executive Director, Pervasive Technology Institute
Associate Dean, Research Technologies
Associate Director, CREST (Center for Research in Extreme Scale
Technologies)
Bradley C. Wheeler
Vice President for Information and Technology & CIO
Professor, IU School of Business
Indiana University - pti.iu.edu
Presented at IBM Multi-customer briefing, Sheraton Hotel, Nov 15,
IEEE/ACM SC11 conference, Seattle, WA
1
2
License Terms
•
•
•
•
Please cite this presentation as: Stewart, C.A. and B.C. Wheeler. 2011.
Cyberinfrastructure Begins at Home. (Presentation) IBM Multi-customer
Briefings, IEEE/ACM SC11 Conference (Seattle, WA, 15 Nov 2011).
Available from: http://hdl.handle.net/2022/13888
Portions of this document that originated from sources outside IU are shown
here and used by permission or under licenses indicated within this document.
Items indicated with a © are under copyright and used here with permission.
Such items may not be reused without permission from the holder of copyright
except where license terms noted on a slide permit reuse.
Except where otherwise noted, the contents of this presentation are copyright
2011 by the Trustees of Indiana University. This content is released under the
Creative Commons Attribution 3.0 Unported license
(http://creativecommons.org/licenses/by/3.0/). This license includes the following
terms: You are free to share – to copy, distribute and transmit the work and to
remix – to adapt the work under the following conditions: attribution – you must
attribute the work in the manner specified by the author or licensor (but not in
any way that suggests that they endorse you or your use of the work). For any
reuse or distribution, you must make clear to others the license terms of this
work.
3
Introduction and Motivation
•
“If you don't know where you are going, any road will get you there” – Lewis
Carroll
•
If you want to go somewhere, set a strategy and pursue it with diligence
•
IU wants to go somewhere:
– To be one of the great public universities of the 21st Century (Michael A.
McRobbie, 18th President of IU)
– To be a leader, “in absolute terms for uses and applications of IT”
(Myles Brand, 16th President of IU)
•
What is cyberinfrastructure, anyway?
– First used in security briefing by Richard Clark in 1998.
– Cyberinfrastructure consists of systems, advanced instruments and data
repositories, visualization environments, and people, all linked together
by software and high performance networks to improve research
productivity and enable breakthroughs not otherwise possible.
4
Outline
•
Some background about Indiana University (fairly extensive)
•
Where we are now
•
How we got here
•
Where we think we are going
•
Generalizations and portable lessons
5
IU – Campuses and Medical School Centers
IU Campuses
IU School of Medicine campuses and clinics
6
IU – Two Core Research/Education Campuses,
Six Regional Campuses
Campus
IUB
IUPUI
IU Northwest
IU South Bend
IU East
IP Fort Wayne
IU Kokomo
IU Southeast
Totals
Academic
appointees
2,973
3,847
393
559
225
838
167
485
8,649
Staff (nonacademic)
5,442
4,793
245
315
152
736
142
250
11,339
Undergraduate
Students
32,367
22,245
5,307
7,790
3,285
12,500
2,978
6,384
80,356
IU community: 121,229 people total
Graduate &
Professiona
l Students
10,097
8,321
662
800
80
800
131
794
20,885
7
Key IU Metrics
IU Budget Category
Unrestricted
2010/2011 Budget
$2,002,537,391
Restricted
$615,974,635
Auxiliary
$407,445,800
Total
$3,025,957,826
IU Health Patient Metrics – 2010/11
Admissions
Outpatient visits
Staffed Beds
115,250
1,882,795
2,889
8
IU’s Goal – Be One of the Great Public Universities of
the 21st Century
As a great public university we have two fundamental missions: education and
research. Indiana University seeks to provide the best possible education to all
of our students, both undergraduate and graduate. It is an education in both
breadth and depth, grounded in both the practical and the transcendent, and
providing a command of the analytical and the expressive. We also seek to
conduct path-breaking research and scholarship of the highest international
standards from the most theoretical to the most applied. …
We are also a public university supported by and with a responsibility to the
citizens of Indiana. They expect us to provide a great education to their sons
and daughters; they expect us to do the best research and scholarship; and
they expect us to be engaged in the life of the State. Therefore, engagement is
a third mission, and it grows out of excellence in education and research.
-IU President Michael A. McRobbie
McRobbie, M.A. 2008. Inaugural address of Michael A. McRobbie: Endurance, Excellence, and the Energy of Change
at Indiana University. http://www.indiana.edu/~pres/speeches/101807.shtml
And … the University Campuses and Culture Are Great
9
http://bloomsofbloomington.org/
http://www.iupui.edu/galleries/272/ndAnnualIUPUIRegatta
Photo by John R. Gentry © Trustees of Indiana University
http://www.sequenza21.com/2011/04/the-premiere-of-bernard-rands%E2%80%99s-vincent-operatic-meditations-on-the-life-ofan-artist/
10
Board of Trustees
President
Michael A. McRobbie
Chief of Staff
Office of the
President
Karen Adams
Vice President for
Research
Jorge José
President,
IU Foundation
Eugene Tempel
Senior
Vice President
and CFO
Neil Theobald
Vice President for
Capital Planning
and Facilities
Thomas Morrison
Vice President for
Public Affairs
and Government
Relations
Michael Sample
Indiana
University
Health2
Executive Vice
President and
Chancellor IUPUI
Charles Bantz
Vice President
for Engagement
William Stephan
Executive Vice
President for
University Regional
Affairs, Planning,
and Policy
John Applegate
Campus
Chancellors
Executive Vice
President and
Provost
Bloomington
Karen Hanson
Vice President for
Information
Technology and CIO
Bradley Wheeler
Methodist
Health
Group, Inc
Vice President for
Diversity, Equity, and
Multicultural Affairs
Edwin Marshall
Vice President and
Director of
Intercollegiate
Athletics
G. Fred Glass
Vice President for
International Affairs
David Zaret
Vice President and
General Counsel1
Dorothy Frapwell
Vice President
University Clinical
Affairs
Craig Brater
Executive Director,
IU Alumni
Association
J Thomas Forbes
Treasurer1
MaryFrances
McCourt
IU East
Nasser Paydar
IU Kokomo
Michael Harris
IU Northwest
William Lowe
IU South Bend
Una Mae Reck
IU Southeast
Sandra PattersonRandles
1Also reports to the Board of Trustees
2A joint effort of Indiana University and Methodist Health Group, Inc.
Revised 10.06.11
Assistant Vice
President and
Executive Director,
Internal Audit1
Michael Gardner
Director,
Affirmative Action
Julie Knost
11
Begin with the End in Mind: End Goals for Service –
IU Researchers, Scholars, Artists, and Students
•
For IU researchers, scholars, students & artists:
– To use information technology within a policy and security framework
that is a model for academia and the US;
– pursue academic and creative activities without limitations created by
resources for data management, computation, or visualization,
accessible with state-of-the-art ease of use;
– examine and present research data, simulation results, or artistic
creations intuitively to enhance scholarly and artistic efforts;
– and have resources available 24 x 365, immediately when desired.
12
13
Begin with the End in Mind: End Goals for Service –
Indiana Residents
•
For Indiana residents:
– To have the benefits of information technology services and information
available from IU;
– the opportunity to obtain a first-class education at all levels; the social
and economic benefits of having IU produce talented graduates, well
educated and ready to pursue interesting and valuable careers;
– a vibrant economy providing satisfying, high-quality jobs with good pay
that provide new career options and entice IU graduates to stay; a high
standard of living and quality of life;
– and an engaging and enjoyable cultural community.
14
Begin with the End in Mind: End Goals for Service –
Others in the US and the World
•
For others in the United States and the world:
– To have the benefits of information technology services and IT-related
information from IU;
– improved quality of life as a result of economic advancement, improved
health, and artistic and literary creations brought about by the
leadership of IU and the state of Indiana.
Goals for 2019
•
IU researchers, scholars, students, and artists will
– Use information technology within a policy and security framework that
serves as a model for academia and the US in general.
– Be able to pursue their academic and creative activities with no limitations
created by access to data, and few limitations caused by access to
computational power.
– Be able to examine and present research results or artistic creations in ways
that are intuitive and enhance effectiveness through
• 2- and 3-D display and interface resources generally available in offices,
labs, and meeting rooms.
• State-of-the-art large-scale facilities located conveniently throughout IU.
– Have access to resources that
• Grow in capability and capacity predictably, steadily, and in ways that
keep IU researchers at the leading edge of discovery.
• Are available resiliently by design (24 x 365 at never less than 75%
aggregate capacity).
• Are available immediately when immediacy is essential.
• Are accessible through interfaces that are intuitively usable by the large
majority of IU researchers.
15
Goals for 2019, con’t
16
Indiana residents will
• Have the benefits of information technology services and IT-related information made available to
state residents by IU.
• Benefit from new, high-quality jobs created by IU’s advanced IT environment (at rates exceeding
the present rate of job creation and contributing strongly to the Indiana economy). Such jobs will
be created in three ways:
– Bring federal money into the state to create new jobs.
– Attract existing companies to locate major business operations in Indiana.
– Create new companies through commercialization of innovations from IU.
• Have available education and training so anyone growing up in Indiana can strive for and obtain
one of these high-quality jobs. The School of Informatics and Computing and other IU schools
matriculate well-educated graduates, many of whom stay in Indiana.
• Have an improved quality of life stemming from these achievements:
• Indiana will rise from 23rd in 2008 to at least 18th in overall rank in the AeA’s annual Cyberstates
report.
• UITS and PTI will aid IU biomedical research and health services communities to improve state
rankings in major health indicators like obesity and tobacco use.
• US and world residents will benefit through access to information technology services and ITrelated information made available by IU; and have improved quality of life, enabled at least in part
by the outcomes of IU discoveries and innovations.
•
•
•
•
Research-oriented Cyberinfrastructure, Informatics,
and Computer Science Units
UITS
– Research Technologies Division
– Networks / GlobalNOC
– Portions of Enterprise Infrastructure
– Portions of Support
– Portions of Learning Technologies
– Portions of Enterprise Software
Academic units
– School of Informatics and Computing
– Maurer School of Law
– College of Arts and Sciences
Pervasive Technology Institute – collaboration involving OVPIT, SOIC,
Maurer School of Law, College of Arts and Sciences, UITS
Key strategies:
– Competition is never within IU, and usually outside the state of Indiana
– Research cyberinfrastructure is a non-zero-sum game from the
standpoint of the university
17
18
CI Facilities
Machine
room
total ft2
ICTC
IUB Data
Center
Avail. ft2
Power
total
Net
power
avail.
Cooling
capacity
total
(tons)
Cooling
capacity
avail.
(tons)
8,300
1,400
600 kW
208 kW
290
150
30,000
15,000
1.46 MW
592 kW
2200
550
•
IU Bloomington Data Center designed to withstand category 5 tornado.
•
Informatics & Communications Technology Complex houses IUPUI offices
and Data Center
•
Innovation Center, dedicated in November 2009, provides a central location
for collaboration between discipline scientists, computer scientists, and staff
through the Pervasive Technology Institute
•
Cyberinfrastructure Building houses OVPIT, UITS staff. To be Gold (or
higher) LEED certified
19
20
21
Computational Systems
Name
Architecture
TFLOPS
Total
Local
RAM (TB) disk (TB)
Big Red
IBM PowerPC 970MP (JS21
blades)
40.96
8
73
Quarry
IBM e1350 Intel Xeon (HS21
blades)
8.96
1.2
5
Mason
HP DL580 G7 Intel Xeon servers
3
8
16
IBM Power5 (p575) database
servers, Intel Xeon E5310
N/A
0.14
50
FutureGrid
IBM e1350
12
6
128
FutureGrid
CrayXT5m
6
1.3
5.5
79.29
15.84
211.7
RDC
Totals
22
Storage Systems
Name
Research File System
Architecture
Disk (PB)
Tape (PB)
OpenAFS
0.06
NA
Data Capacitor
Lustre
0.94
NA
Scholarly Data Archive
HPSS
0.25
5.7
1.23
5.7
Totals
23
An Array of Viz Systems
IUB
IUPUI
Immersive Theaters
Ultra-Resolution
Stereoscopic
24
A Few of IU’s Major Funded Activities
•
XSEDE (Stewart, local PI)
•
FutureGrid (Fox, PI)
•
SEAD (Plale, local PI)
•
Hathi Trust Research Center (Plale, PI)
•
PXGL, ParallaleX, BOOST Graph library (Sterling, Lumsdaine local leads)
•
Indiana CTSI (Barnett, IT lead)
•
Operation Ice Bridge (Knepper, local lead)
•
OGCE (Pierce, PI)
The Local Community is Happy –
Spring 2010 User Satisfaction Survey
25
Average
Satisfaction
Usage
Central research and high performance computers (Big Red,
Quarry, and RDC clusters) [F, Staff, G]
4.16 ± 0.08
92.9 ± 2.2%
14.6%
Center for Statistical and Mathematical Computing (StatMath
Center, [email protected], 855-4724) [All]
4.08 ± 0.06
94.0 ± 1.7%
20.1%
Massive Data Storage Service (MDSS/HPSS) [F, Staff, G]
4.16 ± 0.07
93.5 ± 2.1%
16.2%
Advanced Visualization Laboratory (AVL); www.avl.iu.edu [F,
Staff, G]
4.32 ± 0.07
97.8 ± 1.3%
7.8%
Support for software applications using IU’s high performing
computing resources [F, Staff, G]
4.11 ±0.08
93.3 ± 2.1%
15.2%
3.91 ± 0.08
91.5 ± 2.3%
8.3%
Support for life sciences research (Bioinformatics, The
Center for Computational Cytomics [F, Staff, G]
Overall, how satisfied are you with the UITS
research technology services available at IUB?
Average
How helpful has the information technology
environment at IUB been in your research activities?
Average
4.17 ± 0.06
4.19 ± 0.09
Satisfaction
95.5 ± 1.5%
Satisfaction
95.9 ± 2.2%
Usage
50.0%
Usage
83.9%
26
We’ve Come a Long Way
•
•
•
•
•
In 1997, when IU had its first display at
SCxy, the most common question we
heard was “Why does IU have a booth at
the supercomputing conference?”
So how did we do it?
1998 Indiana University Strategic Plan
(Architecture for the 21st century) adopted
10 Recommendations, 68 Actions
Mostly focused on technology stacks and
the spectra from “competent to excellent”
and “no longer needed to phased out”
27
Year
Notable Events, Grants, and Awards
1955 Marshall Wrubel named first permanent director of Research Computing Center
1964 Frank Prosser, Stan Hagstrom, Steve Young publish FASTRAN compiler
1970s IU develops significant CDC environment, major location for BMDP ports from IB =>
CDC
1981 Bloomington Academic Computing Services formed, first full time administrator (Daniel
W. DeHayes, Business Professor)
1990s IU major VAX facility, focus on computing in humanities, some significant scientific
computing
1997 Michael A. McRobbie comes to IU as first full VP for Information Technology and CIO
1997 vBNS Grant – Connected IU to NSF-funded high speed Internet ($394,000)
1997 First IU display at the international SC conference on Supercomputing
1998 IU IT Strategic Plan adopted
1998 Digital Library IBM SUR (Shared University Research) grant – hardware to support the
Variations digital music library project
1998 GRNOC: Internet2 network engineering and operations support ($3,700,000 contact)
1999 Lilly Endowment, Inc. funds Pervasive Technology Labs ($29M) (Fox recruited)
1999 Indiana General Assembly funds I-Light Network ($5.300,000)
1999 NSF funds network connectivity to Asia (TransPAC) ($10,100,100)
28
Year
Notable Events, Grants, and Awards
2000 NSF funds Variations2 digital music library project ($3,100,000M)
2000 Research @ Indiana Display at SC2000 – first time at this conference that all of the
major research universities of any state coordinate a display promoting their home
state as a whole
2000 Lilly Endowment, Inc. funds Indiana Genomics Initiative ($105,000,000 total;
$6,700,000 for advanced Information Technology)
2001 John-E-Box invented
2001 IU announces first 1 TFLOPS supercomputer owned by a US university; it appears as
the 50th fastest supercomputer on the on Top500 list in the world in November
2001 I-Light activated on December 11
2001 NSF funds AVIDD, first university-owned teraflops supercomputer ($1,800,000)
2001 GRNOC: Indiana Gigapop supporting Higher Education access to national R&E
($500K)
2002 IU establishes first distributed, disaster-resilient tape archive running under the highly
secure HPSS (High Performance Storage Software). Data written to data in tape silos
simultaneously in Indianapolis and Bloomington to ensure reliability of data
preservation
2002 IU joins in life sciences initiative as part of Central Indiana’s life sciences corridor
29
Year
Notable Events, Grants, and Awards
2003 AVIDD achieves more than 1 Trillion calculations per second on standard benchmark
application – the First distributed supercomputer cluster to surpass the 1 TFLOPS
barrier. AVIDD appears in 50th place on Top500 list in June.
2003 NSF funds IP-grid project providing $1,500,000 for IU to become part of TeraGrid, and
$1,400,000 to fund Purdue’s participation
2004 Lilly Endowment, Inc. funds Indiana METACyt Initiative ($53M total; $6.25M for life
sciences IT)
2004 Informatics and Communication Technology Complex Building dedicated on IUPUI
campus – new home for School of Informatics, UITS, and Pervasive Technology Labs
2005 NSF funds IU’s early operations participation in TeraGrid ($440,000)
2005 NSF funds IU as a TeraGrid Resource Partner ($4,100,000)
2005 NSF funds network connectivity to Asia (TransPAC2) ($5.1M)
2005 GRNOC: National Lambda Rail engineering and operations support ($7.5M)
2005 NSF funds Data Capacitor – very fast data storage system ($1,720,000)
2006 Big Red: fastest academic supercomputer in western hemisphere, 23rd fastest
supercomputer in the world on the June Top500 list.
2006 IEDC/IBM grant to double Big Red. Big Red listed as ###___ fastest supercomputer in
the world on Top500 list.
2006 I-Light: First customer connection to backbone ($15.5M)
30
Year
Notable Events, Grants, and Awards
2007 NSF funds PolarGrid ($1.9M) (Fox PI)
2007 NSF funds network connectivity to Pakistan ($950K)
2007 Bandwidth Challenge award – IU moves data across international and national networks
faster than any other supercomputer center competing in this contest at the 2007
Supercomputing Conference
2008 Lilly Endowment, Inc. funds Pervasive Technology Institute ($15M)
2008 I-Light: Network backbone complete ($4.9M)
2009 IU research systems approved by IU Counsel – meets standards for alignment with
HIPAA regulations. IU researchers now able to use IU research systems to analyze
electronic protected health information, thus speeding research
2009 IU Professor Alex Vespignani uses Big Red and I-Light to predict spread of H1N1 virus
and transmit that data to federal agencies in Washington
2009 NSF funds GRNOC: Global Environment for Network Innovations ($330K)
2009 NSF funds FutureGrid ($10.1M) – the fourth of the so-called Track I and II systems
announced officially by the NSF. IBM, Cray major partners
2009- Stewart chairs NSF ACCI Task force on Campus Bridging
2010
2011 IU $4.9M Subcontract as part of XSEDE
2011 NSF funding $1.5M for National Center for Genome Analysis Support
2011 CREST opens – Center for Research in Extreme Scale Technologies
2011 IU announces partnership with Penguin Computing, Inc. for cluster as a service
31
Drivers of Consolidation in the Past (and Ongoing)
•
•
•
Competence and trust
– Summaries of all survey data and every comment (minus obscenities
and identifying references) from 1986 on is available on the web
Facilities and gear
– Especially systems obtained through collaborative grants
Mission orientation
– Constantly asking: what is the mission of IU – are we following it
– We ask the faculty and students to trust our competence…. We trust
their scholarly and artistic leadership
– Serve the home crowd first, the national community when it reinforces
service to the home community. And serve all of the home crowd
– Vendor partnerships. The best partnerships have been based on
complementary strengths and shared goals, and those that persisted
even when there was no money on the table
European Finances and Isaac Newton
32
33
Campus Scholarly Infrastructure
Domain Specific
Discovery & Innovation,
Teaching & Learning
Genomics
Anthropology
Innovation
Innovation
Innovation,
Publication
Arts
Innovation
Visualization
Models
Shared Cyberinfrastructure Line Here?
Physics
Metadata
Visualization
Visualization
Computation
Searching &
Retrieving
Retrieval &
Analysis
Metadata
Metadata
Models
Metadata
Curation
Shared Cyberinfrastructure Line Here?
Necessary
Infrastructure Leveraged
Computation,
Storage
Curation
Curation
Distributed
Storage
Storage
Storage
Networks
Networks
Networks
Networks
Primary
Storage
Discipline Research Stacks…
© Brad Wheeler, Indiana University, Creative Commons Attribution License 2.5
34
Efficiencies and Risk Profile
•
Larger buildings and larger air conditioners more cost-effective per unit
cooling – savings of more than 50% possible
•
Hiding costs by using Graduate Assistants as sysadmins may increase
costs:
– Cost of decreased progress on degree by GA
– How good are most GAs as sysadmins anyway?
•
Cluster in a closet - Cost of down time and underutilization (in exchange for
individual-driven queue policy
•
Network monitoring, electronic security, and physical security
•
Hacker, T.J., and B.C. Wheeler. 2007. Making Research Cyberinfrastructure
a Strategic Choice. EDUCAUSE Quarterly 30 (1): 21-29
35
Money Matters
•
Money is everything (or money can be converted into most anything). Or:
Money is distilled personality
– We know how our money is spent: Activity Based Costing metrics. We
know what our services are, and what the aggregate and units costs of
each service are
•
Our approach has been to build a philosophy of abundance in systems,
people, pay for the people, travel, and opportunities for innovation [No or
minimal internal nuisance fees]
•
Hardware can be a good substitute for money early on if you can get it and
are careful to pay for people to support the hardware and its users
•
Much of IU’s success in many dimensions comes from combining hard-tocombine streams of money
36
Some of the 7 Laws of Money (Michael Phillips,
Shambala Sun)
•
1. Do it! Money will come when you are doing the right thing.
•
3. Money is a dream
•
4. Money is a nightmare
•
6. You can never really receive money as a gift.
•
Why do we (OVPIT/PTI/RT) pursue grant awards?
– In the sincere belief that we are better suited to do some particular task
than anyone else in the world and that the world will benefit from us
doing those tasks
– In the belief that pursuit of a grant award will provide local as well as
national benefits
37
Positive Feedback Loop
Graphic by Bradley C. Wheeler. Used under Creative Commons 3.0 unported attribution license
38
Staff Funding Sources
120
PTI Soft
PTI Base
100
RT Soft
FTE
80
RT Base
60
40
20
96 97 98 99 00 01 02 03 04 05 06 07 08 09 10
Fiscal Year
39
Evolution of IT Services
2008 – IU IT Strategic Plan
•15 Recommendations
• 15 Recommendations
• 72 Action Items
• Role and function based, not
technology based
•72 Action Items
Edge
Trust
?
Leverage
Trust
?
Edge
40
Empowering People – Recommendations Relevant to
Research
•
Recommendation 1: Indiana University's national and international
leadership should be sustained through continued maintenance and
advancement of an IT infrastructure that is supported by sound fiscal
planning.
•
Recommendation 9: Indiana University should provision appropriate "data
utilities" for administrative data/information, research data, teaching and
learning resources, and multimedia scholarly life. These utilities should
provide convenient, timely, and secure access to university data/information
by the IU community and authorized collaborators beyond IU.
•
Recommendation 15: While Indiana University should advance IT-enabled
research across all disciplines, it should also focus on a few highly
promising opportunities for which it has a skills, knowledge, and reputational
advantage to push the frontiers of IT-enabled research and scholarship.
41
Factors Driving Consolidation
within the University in the Future
•
Economies of scale in management
– Air conditioning
– Electricity supply
– Staff
•
Security
•
VM hosting – intelligent
infrastructure
•
Economies of vendor partnerships
and large purchases
•
Redundant backups (2 locations)
and our own network
42
Intelligent Infrastructure
•
Virtual systems supply the infrastructure and network capacity necessary to
host your applications, while optional disk storage on UITS enterprise-class
SANs (storage area networks) ensures your files are extremely secure and
always available.
•
Backup solutions provide cross-site backups and cross-campus failover
options, which isolate you from potential disasters by securing your backup
files within hardened data centers.
•
This is a charge-back solution - $425 / system / year; $1 /GB / year
•
Adoption and growth have been phenomenal – service introduced in July
2007, set up 1000th VM in 2009, today have a total of 1981.
•
Tremendous energy efficiencies - 26 VMs per physical core
•
Departmental staff are able to focus more on things that are unique to their
departments – and with implementation of CMS(es), IU’s web presence is
improving
43
IUanyWARE – Core Attributes
•
•
•
•
Commodity delivery of IU applications (> 200)
Target Community: IU Staff, Students and Faculty
Generic easy to consume features
OS agnostic
• Android
• Chrome
• iOS
• Windows
• Linux
• Access to IU software anywhere anytime and on any device for 100,000
students and over 17,000 full-time staff.
• Allows virtual consolidation of service delivery without physical consolidation
of students in student computing clusters!
44
PTI - History and Goals
•
•
•
•
Started by $30M grant from the Lilly Endowment to create the Pervasive
Technology Labs and fuel growth of the IU School of Informatics and
Computing. 2nd round funding of $15M to create Pervasive Technology
Institute in 2008
Overall:
– Invent, develop, deploy, deliver
– Serve IU’s flagship ‘brand’ re this activity in advanced IT
PTI Research Centers Goal:
– Research Excellence
PTI Service & Cyberinfrastructure Centers Goals:
– Support Development, deployment, support of research IT
– Support research-related activities by IU community generally, PTI
Research Centers particularly, and national research community when
in IU’s interests
– Aid state of Indiana
• Baseline level of workforce & economic development activities
45
PTI Structure
•
PTI is a collaboration among School of Informatics and Computing; Maurer
School of Law; Office of the Vice President for Information Technology;
University Information Technology Services
•
Consists of two types of centers:
– Research Centers: D2I; CACR; DSC; CREST. Needs and
communication coordinated by Managing Director.
– Service/Infrastructure Centers: Research Technologies Division of UITS.
Needs and communications coordinated by Executive Director
•
Managing Director and Executive Director report to VP for IT and are jointly
responsible for success of PTI
•
Executive Committee: Relevant Deans, VP for Engagement, VP for IT
•
Reports administratively to OVPIT
46
Economic Development
•
•
437 full time job-years of employment created within IU by grants brought
into Indiana since 1999
Based on economic multipliers, 2,705 full time job-years of employment in
the state of Indiana since the start of PTL in 1999
2002
2004
2008
2010
Indiana
30
29
33
28
Illinois
19
21
21
20
Kentucky
46
48
47
47
Michigan
24
25
26
26
Ohio
27
24
36
29
Milken Institute’s State Technology and Science
Index for Indiana and bordering states
47
Cyberinfrastructure Begins at Home,
But it Doesn’t End There
Some CI Resources Available to Science and
Engineering Researchers in the US (March 2011)
NSF Track 1
Track 2 and other major facilities
Campus HPC/ Tier 3 systems
Workstations at Carnegie research
universities
Volunteer computing
Commercial cloud (Iaas and Paas)
0
2,000
4,000
6,000 8,000 10,000 12,000
TFLOPS
Based on: Welch, V.; Sheppard, R.; Lingwall, M.J.; Stewart, C. A. 2011. Current structure and past history of US
cyberinfrastructure (data set and figures). http://hdl.handle.net/2022/13136
48
Adequacy of Research CI
Never (10.6%)
Some of the
time (20.2%)
Most of the time
(40.2%)
All of the time
(29%)
Responses to question asking if researchers had sufficient access to
cyberinfrastructure resources – survey sent to 5,000 researchers selected
randomly from 34,623 researchers funded by NSF as Principal Investigators 20052009; Results based on 1,028 responses.
Stewart, C.A., D.S. Katz, D.L. Hart, D. Lantrip, D.S. McCaulay and R.L. Moore. Technical Report: Survey of
cyberinfrastructure needs and interests of NSF-funded principal investigators. 2011. http://hdl.handle.net/2022/9917
49
50
Local and National: Use Facilitated
•
XSEDE – 11 Projects, > 4M
TeraGrid normalized units / year
•
Open Science Grid
•
~$1M/year in resources for IU
researchers
51
Clouds Look Serene Enough
Photo by http://www.flickr.com/photos/mnsc/
http://www.flickr.com/photos/mnsc/2768391365/sizes/z/in/photostream/
http://creativecommons.org/licenses/by/2.0/
52
But is Ignorance Bliss?
•
In the cloud, do you know:
– Where your data are?
– What laws prevail over the physical location of your data?
– What license you really agreed to?
– What is the security (electronic / physical) around your data?
– And how exactly do you get to that cloud, or get things out of it?
– How secure your provider is financially? (The fact that something seems
unimaginable, like cloud provider such-and-such cloud provider going
out of business abruptly, does not mean it is impossible!)
•
NB: Thomas Friedman wrongly insulted the state of Indiana
53
There Are Good Definitions of Cloud Computing - NIST
•
Cloud computing is a model for enabling convenient, on-demand network
access to a shared pool of configurable computing resources (e.g., networks,
servers, storage, applications, and services) that can be rapidly provisioned
and released with minimal management effort or service provider interaction.
•
This cloud model promotes availability and is composed of five essential
characteristics (On-demand self-service, Broad network access, Resource
pooling, Rapid elasticity, Measured Service); three service models (Cloud
Software as a Service (SaaS), Cloud Platform as a Service (PaaS), Cloud
Infrastructure as a Service (IaaS)); and, four deployment models (Private
cloud, Community cloud, Public cloud, Hybrid cloud).
•
Key enabling technologies include: (1) fast wide-area networks, (2) powerful,
inexpensive server computers, and (3) high-performance virtualization for
commodity hardware.
•
http://www.nist.gov/itl/cloud/index.cfm
54
Above Campus Services
•
•
•
•
•
"We are seeing the early emergence of a meta-university — a transcendent,
accessible, empowering, dynamic, communally constructed framework of
open materials and platforms on which much of higher education worldwide
can be constructed or enhanced.” Charles Vest, president emeritus of MIT,
2006
“Shaping the Promise of Cloud Computing for Higher Education” – Wheeler
and Waggener
Goal: achieve economy of scale and retain reasonable measure of control
Examples:
– Internet2 and HP
– IU, UM, UVa, Penguin Computing Inc.
See: Brad Wheeler and Shelton Waggener. 2009. Above-Campus Services:
Shaping the Promise of Cloud Computing for Higher Education.
EDUCAUSE Review, vol. 44, no. 6 (November/December 2009): 52-67.
http://www.educause.edu/EDUCAUSE+Review/EDUCAUSEReviewMagazineVolume44/AboveCa
mpusServicesShapingtheP/185222
55
Campus Bridging
In early 2009 National Science
Foundation’s (NSF) Advisory
Committee for Cyberinfrastructure
(ACCI) charged six different task
forces to make strategic
recommendations to the NSF in
strategic areas of cyberinfrastructure:
•
•
•
•
•
•
Data
Grand Challenges and Virtual
Organizations
High Performance Computing
Software and Tools
Work Force Development
Campus Bridging
56
Highlights of Campus Bridging Taskforce Report
Conclusions
•
Stop the madness – just adopt InCommon /
XSAML certificates
•
NSF must lead the community in
establishing a blueprint for a National CI
•
Every institution of higher education should
have a strategic plan for IT
•
The NSF should fund activities that support
the evolution and maturation of
cyberinfrastructure through careful analyses
of needs (in advance of creating new CI
facilities) and outcomes (during and after
the use of CI facilities)
•
pti.iu.edu/campusbridging
57
NCGAS and Campus Bridging
•
The National Center for Genome Analysis Support
•
A Cyberinfrastructure Service Center affiliated with the Pervasive
Technology Institute at Indiana University (http://pti.iu.edu)
•
Dedicated to supporting life science researchers who need computational
support for genomics analysis
•
Initially funded by the National Science Foundation Advances in Biological
Informatics (ABI) program, grant # 1062432
•
Provides access to genomics analysis software on supercomputers
customized for genomics studies including IU Mason system, POD IU,
Stampede, Gordon, and other XSEDE systems
•
Particularly focused on supporting genome assembly codes such as:
– de Bruijn graph methods: SOAPdeNovo, Velvet, ABySS,
– consensus methods: Celera, Newbler, Arachne 2
•
For more information, see http://ncgas.org
Some Things We Really Got Right from the Start
•
Key assets for any university: faculty
•
Key assets for the IT organization: reputation for good character, people, then
everything else
•
It takes a lot of Ph.D.s, and a mix of domain scientists and computer scientists
•
Put health and family first, no matter what, at least in times of crisis
•
Mix of funding sources
•
Publish, even if we wouldn’t perish otherwise
•
Participate in community, partner effectively
•
Engage faculty as partners and leaders
•
Timing, some days (SC, challenges, Top500)
•
Truth is a wonderful thing; no cynicism zone
•
Lots of feedback – difference between input rights and decision rights (and
sometimes input rights don’t exist)
•
Understanding how it is we do economic development
58
59
And Some Things We Didn’t Get Right to Begin With
•
Leading where faculty aren’t
•
FIFO
•
Believing we had to put up with staff who didn’t want to be part of our team
because they said they were unique and we had to
•
Reading Sun Tzu without Lao Tzu, and vice versa
•
Believing too much that if we paid attention to 4-space, press-space would
take care of itself
•
Listening to funding agencies too much
•
Work overload, insufficient attention to life balance and creativity
•
If I (Stewart) had one thing to do over … we would have done fewer projects
each at larger scale early on
60
And Things We Think We Have Learned
and Now Internalized
•
Play to win, or not at all
•
The community is too good for any one group to win very often, unless you’re
very very good and have lots of money to pitch in
•
Being good is often not enough
•
One time money is too easy to get without base
•
Heroic effort is not a sustainable strategy
•
Partnering and openness
•
Don’t hone in on bad ideas too quickly (but be careful of the side effects in terms
of project thrash)
61
Key General Lessons
62
Leadership Matters
Myles Brand
Indiana University
President 1994-2002
Michael A. McRobbie
Indiana University
VP for IT & CIO 1997-2007
President 2007-
… and teamwork matters too!
63
Assertions about CI at Home and Beyond
1) Well-tuned campus CI enables greater research outcomes
2) Well-tuned campus CI is very expensive …though less expensive than other
models
3) Well-tuned CI is possible only in the presence of clear strategy at the institutional
level
4) Campus CI is (mostly) a coordination problem in the guise of a money problem
5) Above campus CI is not magic, and does not take away responsibility for
delivering service to local CI users… but properly employed can be a useful
strategic tool
6) Leadership in service of mission is the essential means of creating sustainable
campus CI
7) Effective partnership is essential – colleagues, community, institutions, vendors
If I Were Starting from Scratch or Nearly So…
•
•
•
•
•
•
•
•
Find some faculty advocates and cultivate relationships with them
If you are at a place with good leadership => then educate leadership, build
consensus for some small pilot projects among faculty
If you are at a place without good top level leadership (and that happens)
=> then work directly with faculty to build consensus for some pilot projects
Build trust
Take some risks. If the organization you work for won’t support you when
you take reasoned risks and your only success is a learning experience,
you deserve a better organization as your employer.
If you don’t have anyone on your staff with a Ph.D., find someone with a
Ph.D. (not in CS) and a good customer service attitude and hire them.
Assign one person the full time job of supporting people in use of XSEDE
and Open Science Grid – particularly the process of applying to XSEDE
Pay close attention to public relations, but lead with the scholarly results
and why they matter, not where you are on the Top500 list
64
65
Get Some Facts and Have Fact-based Discussions
•
•
•
•
•
•
•
•
Pay someone else to do a real, human-subjects approved survey of your
clients and repeat it every year
Initiate a service definition and Activity Based Costing exercise
Publish, including effective self-publishing:
– NSF-format scholarly nuggets
– Technical reports
– Use the nearest real digital repository and a print on demand service
(we use CreateSpace)
Communicate concisely, well, and in English. Hemingway is easier to
process than Faulkner
Make the economic and workforce case to your community and state
Remember highlights of your history and repeat them often
Remember lowlights of your history, study and understand then, remember
the lessons for you and your organization, but don’t proactively remind the
people you serve about them
Study history of leadership, contention, and conflict
Work This Cycle – Even If You Just Start With a Small
Hardware Grant-in-Kind
NB: Of the things we have done, the cyberinfrastructure service over the years has
been the easier. The development of PTL and PTI has been by far the harder. Some
Institutions that have not done well have started with the research center first, before
Establishing a base of trust and good will with general service to the community.
Graphic by Bradley C. Wheeler. Used under Creative Commons 3.0 unported attribution license
66
Participate in Community
•
•
•
•
The definitions of friends and colleagues are different. An organization is
effective when colleagues collaborate effectively.
Remember who your competition is (and isn’t)
Collaborate well, and maintain strong partnerships with other academic
organizations and vendor(s).
Participate in community
– Volunteer for conferences, serve on review panels
– Join appropriate groups (CASC perhaps)
– Submit benchmarks to SPEC, HPCC, Graph500, Green500
– Join InCommon or at least set up a mechanism to deliver InCommonapproved XSAML credentials when needed
– Get someone to become an XSEDE Campus Champion (preferably
outside IT shop)
– Make your materials reusable under a sensible license (Creative
Commons for docs / some reasonable public license for data)
– Remember credit is a nonconserved property
67
Alternate Financial Models
•
•
•
•
•
IU operates on Responsibility Center Management with a non-optional
financial allocation for certain central services … including IT
IU has also benefitted greatly from support of the Lilly Endowment
The relationship models IU has developed are likely fairly portable
Financial models at different universities vary
If you want to look at different financial and organizational models, I
suggest:
– Purdue – the best current example of a community funding model (Penn
State was the pioneer in this area)
– Texas Advanced Computing Center – they do things big in Texas, but
the model of excellent science and strong relations to alums and donors
might be more transportable than one would imagine at first blush
– The work that Amy Apon did at University of Arkansas bears note as
well – basically building on expertise and a Major Research
Infrastructure award from the NSF to bootstrap a small and very good
center into existence
68
69
And Remember That We Are Living in the Myth of Sisyphus
"The struggle itself...is
enough to fill a [person’s]
heart. One must imagine
Sisyphus happy.”
–Albert Camus
Sisyphys (1548-1549) by Titian, Prado Museum, Madrid, Spain
http://en.wikipedia.org/wiki/File:Punishment_sisyph.jpg
This work is in the public domain in the United States, and those
countries with a copyright term of life of the author plus 100 years or
fewer.
70
Absolutely Shameless Plugs
•
•
•
•
XSEDE12: Bridging from the eXtreme to the campus and beyond
July 16-20, 2012 | Chicago
The XSEDE12 Conference will be held at the beautiful Intercontinental
Chicago (Magnificent Mile) at 505 N. Michigan Ave. The hotel is in the heart
of Chicago's most interesting tourist destinations and best shopping.
Watch for Calls for Participation – coming early January
•
And please visit the XSEDE and IU displays in the SC11 Exhibition Hallway!
Thanks
•
•
•
•
•
•
•
•
IBM for a long, and VERY productive, collaboration
All of the IU Research Technologies and Pervasive Technology Institute staff
who have contributed to the development of IU’s advanced cyberinfrastructure
and its support
All of our faculty collaborators, especially the PTI-affiliated faculty
Thanks to the colleagues who have contributed directly to this presentation:
Malinda Lingwall, Vince Cannon, Greg Moore, Rob Lowden, Duane Schau, Eric
Wernert, Mike Boyles, Bill Sherman
Those involved in campus bridging activities: Guy Almes, Von Welch, Patrick
Dreher, Jim Pepin, Dave Jent, Stan Ahalt, Bill Barnett, Therese Miller, Malinda
Lingwall, Maria Morris, Gabrielle Allen, Jennifer Schopf, Ed Seidel
NSF for funding support (Awards 040777, 1059812, 0948142, 1002526,
0829462, 1062432, OCI-1053575 – which supports the Extreme Science and
Engineering Discovery Environment)
Lilly Endowment, Inc. and the Indiana University Pervasive Technology Institute
Any opinions presented here are those of the presenter and do not necessarily
represent the opinions of the National Science Foundation or any other funding
agencies
71