spring 1999 internet2 members meeting 28 april 1999
TRANSCRIPT
New Members since Fall 98Regular University Members
Baylor College of Medicine Brigham Young University College of William and Mary Drexel University East Carolina University New Jersey Institute of Technology
New Members since Fall 98Regular University Members
Portland State University South Dakota School of Mines South Dakota State University Stephen F. Austin State University Tufts University University of Alabama - Huntsville
New Members since Fall 98Regular University Members
University of California - Riverside University of Maine University of Memphis University of Montana University of North Texas University of South Dakota
New Members since Fall 98Regular University Members
University of Texas Southwestern Medical Center at Dallas
Washington University Wayne State University Wright State University
New Members since Fall 98Affiliate Members
Association of Universities for Research in Astronomy (AURA)
Bradley University LaNet (Louisiana) Survivors of the Shoah Visual History
Foundation University of North Carolina General
Administration
New Members since Fall 98Affiliate Members with Collaboration Site Status
Howard Hughes Medical Institute Jet Propulsion Labs The National Institutes of Health
Beyond the BackboneConnecting Ideas, Priorities, People and Resources
Ron Johnson
Vice President and Vice Provost
University of Washington
The national backbone, as hard as that is to do, it is the easy part!
2nd easiest are the campus infrastructure issues
Then there’s organizing the GigaPoP
Connecting the Ideas, Priorities, People and Resources
Then there’s the GigaPoP(s)
Working with:
RBOC’s etc (and the local loops we ‘regionals’ used to have, but now are gone :-(
NSF (where network funding mechanisms and great programs are still held hostage by lawyers and inspector generals)
Gigapops ...
the nearest Qwest pop is >1000 miles away
Or there are a few too many Qwest/MCI et al etc. pop’s on the block :-)
Organizational issues
Understanding & Connecting with Mission Agency NGI efforts
Priming the local, regional and national Applications and Content Pumps
Soliciting/packaging Proposals -- PI’s with meritorious apps
Delivering on the Proposals & Initiatives
Getting the word out & getting mindshare
$$$, people with clue
But what are we actually doing? (other than going to I2 mtgs?:)
Created Core Pacific/Northwest GigaPoP & SNNAP etc. with peering and internet aggregation points
Multiple local gigabit attachments, city fiber grid -- w/I2 stepped up connections - NOAA et al (& attached vBNS & Abilene)
Campus Infrastructure Upgrades -- course corrections to leverage opportunities
Delivering on our related Proposals & Commitments
(RE)-Creating ‘regional’(!) network structure and connectivity
Wrestling with the ‘Region’s’
L O N G local loops (3k to Fairbanks - - Bozeman, Moscow etc.)
Commodity Internet Integration & Aggregation (esp. in areas w/o competitive local options) needs
Org. and Relationships building
OBSERVATION:
PERSISTENT REGIONAL HUB INFRASTRUCTURE - It’s hard to believe that it is far harder to create or resurrect a persistent R&E infrastructure than it is to (re)create a national backbone -- but it is!
OBSERVATION
WE need to do a far better job of developing a broad understanding within the ‘beltway’, and among our good PACI etc. friends that this is the case, and we need to seek their help in accomplishing that.
Working with mission agency NGI networks, pop’s, connections
Sorting out whether some other agencies (e.g. NIH) will play active roles, and how some other agencies will be involved (esp. when their ‘branches’ &/or PI’s are on our doorsteps)
OBSERVATION
We need to do a better job (for we sure haven’t succeeded thus far) of working to encourage NGI “mission” agencies to follow the promising examples that folks like Leighton of ESNET are setting, that exist as models to develop joint plans that really will effectively encompass university researchers, and which will leverage regional hubs
OBSERVATION:
Meanwhile NSF CISE really does need to be the vehicle for managing & distributing IT2, post-’vBNS’ , and the bulk of NGI resources
Back on campus:
Getting the word out (soon enuf but not too soon, wide enuf but not too wide, loud enuf …
Applications/Content/Science/Arts … marshalling efforts - e.g. ResearchTV
Reshaping Efforts/Directions - middleware, contemplating trust fabric in a larger context
Saying “no” to non-RU1’s, community colleges, 4 years, k-12. Museums, galleries, libraries, other ISP’s?
Sorting through international connections
E-Ticket -- Looking to an exceedingly interesting future
Campus Infrastructure Investment
Continuous Investment is Required Applications Drive Technology and
Resources Retention of Expertise is a Challenge Risk and Return On Investment
• Costs are insignificant when considered as a function of the competitive importance to the institution of advanced communications infrastructure.
Rapid Evolution
Technology generation period is about 18 months.
Continuous refreshment enables us to take advantage of economies of scale and of scope.
1977
$4.2M annually< 4000 connections$1,050 /connection
1999
$13M annually> 40,000 connections
$300 / connection($116 1977 dollars)
Virginia Tech Network Capacity and Price/Performance
64Kbps10Mbps
100 Mbps
400 Gbps
622 Mbps
1978 1982 1986 1990 1994 1998
Network Capacity(non-linear)
2002
CNS Auxiliary
IntegratedVoice/Data
EngineeringStudent
Computers
Satellite-basedDistance Learning
BEV
NET.WORK.VIRGINIA
Internet2
Residence Halls Ethernet
CyberschoolFDI
InternetE-mail
VP Info SystemsComputers All
Freshmen
LMDSDevelconDataswitch
LocalNet
CBXEthernet
FDDI
ATM
DWDM
Cost per Megabit
$ / Mb
$10,000
$0
Capacity
Mile
ston
es
Cisco7500
IPRouter
622 Mbps ATM
155 Mbps ATM
Distribution
Main Distribution
Frame
10/100 Switch
10/100 Switch
10/100 Switch
10/100 Switch
ded
ica
ted
1 G
bit/
s d
own
links
Type II/III Building
Type I Building
10/100 Switch
ATM attached desktop
ATM attached server
Videoconference sys
Catalyst5000
Catalyst5000
ASX-200WGATM Switch
ASX1000ATM
Switch
Virginia Tech ATM Core
BURASX1000
ISBASX1000
HILASX1000
OWEASX1000
CASASX1000
SHAASX
200BX
622 Mbit/s155 Mbit/s
Main DistributionFrames
Virginia Tech IP Core
MCI204.70.78.20/30
128.173.100.0/24 128.173.101.0/24
vBNS
ESnet
Internet
OC3Abilene
DS3OC3
245 Mbps
BUR7507
ISB7507
1
HIL7507
ISB7507
2
CAS7507
Net.Work.Virginia
Beyond The Backbone
Bonnie Neas, Director
Information Technology Services
North Dakota State University
and
North Dakota University System
Program Committee MeetingFall 1999 Member Meeting
Noon - 1 pm Meeting Room 14 (upstairs) pick up a box lunch first from Renaissance
Ballroom West
Concurrent Breakout Sessions
Up one level (Meeting Room level) 2:15 - 3:30 pm
Followup discussion with NSF I2-DV discussion I2-DSI discussion GigaPoP presentation
Concurrent Breakout Sessions
Up one level (Meeting Room level) 4:00 - 5:30 pm
Middleware discussion International MoU partners reports NanoManipulator SURA BOF meeting
Gala Reception tonight!
Air and Space Museum buses will run continuously between
Renaissance and Museum first bus 7:15 pm last bus 10:15 pm thanks to AT&T for their generous
sponsorship!
Discussion Outline
What is Middleware• why is it important • why is it hard
What are the major components of middleware
Where are we today Internet2 Middleware Initiative
What is middleware? specialized networked services that are shared by
applications and users a set of core software components that permit scaling
of applications and networks tools that take the complexity out of application
integration sits above the network as the second layer of the IT
infrastructure the intersection of what networks designers and
applications developers each do not want to do
Why is middleware important?
Enables a scalable approach to ubiquitous computing
With interoperability, permits academic research community to collaborate nationally and internationally and utilize specialized network resources
Middleware is necessary for the deployment and creation of advanced applications
As an end user it offers me...
single logons a single place for my personalized
information institutional file systems for archive and
backup permission to use distant scientific
resources access to digital libraries
As a developer it offers me…
Building blocks for faster software development
Abstractions to hide networking details
The ability to focus on content issues rather than programming challenges
As a CIO it offers me...
the second layer of the IT infrastructure leverage to both instructional and
research missions scalable infrastructure for applications a more secure IT environment a chance to resolve the silos and
fiefdoms of information
A Map of Middlewareland
Network-layer middleware
Core middleware
UbiquitousComputingUpperware
ResearchOriented
Upperware
BusinessUpperware
Core Middleware
Identity - the first characteristics of who you are Authentication - how you prove or establish that
you are that identity each time you connect Directories - where the rest of an identity’s
characteristics are kept Authorization - what an identity
is permitted to do
Middleware and Higher Ed
We have different characteristics than corporations• mobility multiple roles per individual • non-monetary economies Open Records and FERPA
Some pieces have small commercial markets• remote instrument control coscheduling of distributed
resources • digital libraries
By role and mission we are often pre-market• roaming profiles interoperability among competitors
What makes it hard?
Technology issues• where are the products• Where are the protocols for interoperability
Practice issues• Design and designers• What’s the metadata• Top-level authorities
Policy issues• FERPA and Open Records• Legacy systems, legacy thoughts
I2MI
Starting discussion - NPPAC and TAC Next steps Doing the work - on a continuing
basis Dissemination - dissemination and
dissemination
NPPAC Discussion
Defining the IT environmental needs Recognizing current constraints on
campus IT organizations Policy issues and reality checks
Technical Advisory Committee
Defining the technical requirements University IT architects - the best and the
brightest• CMU, Stanford, MIT, Michigan,
Washington, California, Brown, Notre Dame
Partners - Educause, PACI, CNI I2 Staff
I2MI Goals
Interoperable middleware infrastructures• help campuses develop infrastructures• coordinate interoperability
Get some key apps involved• digital video• inter-institutional resource sharing
Work for the PACI and other research needs Link to network layer needs Understand what automating the policy layer takes
First Phase Work (6-12 mo)
Capture what we know now. Get campuses understanding the issues. Establish some basic public domain tools. Synchronize activities with partners. Influence vendors. Engage and inform political processes -
AACRAO,NACUBO, NACUA, NASULGC,etc. Continuously distill and disseminate.
Intelligence gathering
Talk to advanced campuses • Identify common issues• Identify distinctive factors• Gather implementation approaches• Leave in place ongoing intelligence gathering
Work with PACI sites Digital video Digital libraries and inter-institutional resource
sharing Help distill and disseminate vendor information
Dissemination
Interoperability requires broad information and consensus
Roadmaps Work with our partners Work with the membership
• In seminars • In tech talks on the net • On the web• At member meetings • In print
I2MI RoadMaps
Outline the issues Suggest decision criteria for
implementations Identify vendor products and
associated issues Point to leadership approaches Establish interoperability criteria
Establish tools
Vendor interoperability Link middleware to applications Foster public tools
• Solidify key glues and adhesives• Point to reference implementations
Create standards and consensus
Synchronize partners Convene consensus processes on
needed conventions Educate political processes -
AACRAO,NACUBO, NACUA, NASULGC
Second Phase Work (12-24 mo)
Create cadres of institutions to somewhat synchronously deploy campus-wide middleware.
Provide technical support services and facilitate intercampus communications.
Harvest information and disseminate.
Supporting second phase participants
Communicate and enhance participant efforts Identify key technologies and associated products Foster middleware-based applications Bring in expertise Capture design wisdom and the art of the practice Influence vendors to make products more
higher-ed friendly Inform campus policy processes