Bierring_AdvNetDay - University of Victoria

courageouscellistAI and Robotics

Oct 29, 2013 (3 years and 7 months ago)

76 views







CANARIE, CA*net 4 and


Intelligent
-
Infrastructure


University of Victoria

November 28, 2005



Andrew K. Bjerring

President and CEO

Outline


Background


CA*net 4, Lightpaths and UCLP


Grids and Intelligent Infrastructure



Background


The Ultimate Driver...

...and the effect on networks

“When the network is as fast as the computer’s internal links, the machine
disintegrates across the net into a set of special purpose appliances”
-
George Gilder

The Evolution of the Internet



First wave


Application:

Transfer of text and data


Network:

An add
-
on to the computer


Organizations:

Real and local

Second wave


Application:

Widespread access to images, sound and video.


Network:

A partner with the computer


Organizations:

Increasing “virtual” elements

Third wave

*


Application:

Linking services; real time transactions


Network:

Integrated into a



“cyber infrastructure” or “intelligent infrastructure”


Organizations:

Emphasis shifting to
“communities of interest”



*Forbes.com: “...the next phase of [the networking} revolution comes from adding new intelligence and services to the electro
nic

"plumbing" that now
envelopes the very way we conduct business.

From the Web to wireless, a new level of
intelligent infrastructure

is fast
-
changing the business landscape”.
(April, 2005)



CANARIE’s First 12 Years

(1993
-
2005)



Advanced Networks:

4 generations of CA*net and associated regional
nets



Next
-
Generation Applications:
Over 200 R&D projects focusing on e
-
learning, e
-
health, e
-
business, e
-
content and underlying technology



Domestic and International Outreach:
Broad range of more general
awareness raising, community development and branding activities.

Next Generation Applications


Phase 1: 1993
-
1995

$26 m


Stand
-
alone projects



Phase 2: 1995
-
1999

$80 m


Collaborative projects



Phase 3: 1999
-
2004

$78 m


Strategic sectoral development



Phase 4: ??


Intelligent infrastructure and e
-
solutions

Example of Sectoral Development:

E
CCnet



Industry
-
wide data synchronization solution for supply chain
management in grocery retail, foodservice, and pharmacy sectors


Large enterprises, MNEs and SMEs all represented


For grocery retail sector: 2,500 companies, 300,000 products,
85% of transactions


$4.5M CANARIE funding; over $2.0M paid back to date



General Lesson Learned

Next
-
Generation applications involve applying new technologies to
develop and re
-
design processes and relationships
across
organizational boundaries





Technological Innovation



Web services and grids


Convergence of networks and IT


Leading to shared “intelligent
infrastructure”

Process and Relational Innovation



Opportunities for productivity gains


Inter
-
organizational processes


Leading to a “community of interest”
(COI) based planning

Status of CANARIE Strategy

(2005
-
2010)

CA*net 4



Current funding ends March 31, 2007


Extension to 2010 approved in principle by federal cabinet


Funding for post March 31, 2007 pending


Change from 3
-
5 year funding model part of proposal


“E
-
solutions” Proposal


Element of Industry Canada’s broader ICT Sector Strategy


Election call has introduced further delays and uncertainties

CA*net 4, Lightpaths
and UCLP

CA*net Generations

CA*net 4 Overview



National Backbone:
Connects regional R&E networks (ORANs) in
every province, with links to global counterparts



Very High Bandwidth:

Currently based on five 10 Gbps wavelengths



Hybrid Design
:

Fraction of one lambda is run as layer 3 network;
balance of capacity is allocated as separate “lightpaths” to specific
research projects and services (a.k.a. “lambda network”)



Enables e
-
science:
For example, supports dedicated connections to:
beamlines at Canadian Light Source; remote HPC and visualization
clusters; telescopes; distributed sensor arrays; data repositories; etc.

Global Lambda Integrated Facility

Research & Education Network bandwidth for scheduled application and middleware research experiments

LightPath Examples

1.
KREONet2 to StarLight


Network link



10G


2.
UoAlberta to CERN


ATLAS High Level Trigger


1G

3.
Montreal to Emily Carr Institute

Art Net




155M

4.
McGill to Fermilab


Data Transfer



1G

5.
Ireland to Chicago


Network link



1G

6.
Victoria to Ottawa to Montréal

Linking NRC Institutes


1G

7.
Nortel (Ottawa) to Chicago

Research



1G

8.
Alberta (WestGrid) to Fermilab

Data transfer



1G

9.
TRIUMPH to CERN


Data transfer



10G

10.
Chicago to NewYork


GLORIAD



1G

11.
Montreal to Hull


RISQ backup



1G

12.
Kingston to Chicago


Data Mining



1G

13.
Ottawa (CRC) to Barcelona

UCLP Demo



1G

14.
California to Europe


Caltech performance testing

1G

15.
Seattle to Ireland


UCLP demo



1G



CA*net 4 Architecture

Calgary

Regina

Winnipeg

Montreal

Toronto

Halifax

St. John’s

Charlottetown

Chicago

Seattle

New York

Saskatoon

Victoria

Vancouver

Edmonton

Ottawa

Fredericton

Optical
Switches

User Controlled LightPaths

(UCLP)

Status



Software to partition and distribute control of optical switches


Allows users to create and manage end
-
to
-
end lightpaths from the edge


Uses web
-
services based Service
-
Oriented
-
Architecture (SOA)


Three designs for initial version (UCLP v1) funded


Diffusion is underway (open source)


Deployed across CA*net 4, KREOnet, I2Cat


Used by ORANO, Netera, BCnet, HEAnet, RISQ


Being tested by Internet2, SURFnet, AARnet, Nortel, Cisco, Meriton and others


UCLP v2 development underway with three additional contracts


New CA*net 4 Programs

ORAN Support and Outreach:



$15 million


to support lightpath capability


to ensure universities, colleges and remote campuses can access CA*net 4


to undertake promotion and awareness activities ($1 m)


provincial allocations confirmed

STATUS:

Several Statements of Work approved and work started


CANARIE Connections Program:



$10 million


to assist federal labs and private sector to connect sites and facilities to CA*net 4


Submissions received from NRC, NRCan, Environment, DFO, Agriculture Canada, Stats Canada
and others

STATUS:

First connections approved; first departmental MOU signed; other discussions continuing


Intelligent Infrastructure:



$15 million


to extend Service Oriented Architecture (SOA) to instruments, sensors and controllers and other
next
-
generation applications

STATUS:

First seven projects chosen; other discussions re “gaps and opportunities” underway

Grids and Intelligent Infrastructure

ATLAS Grid


TRIUMF has been selected as Tier 1 site for CERN data


One of only ten Tier 1 sites in the world


Requires 10 Gbps connection to CERN by Q1
-
2006 and 5 Gbps
connection to backup site


Will also require dedicated 2 Gbe lightpaths to Tier 2 sites


Circuit may also be used to support a routed network for HEPnet
(Canada’s High Energy Physics network)

Astronomical Data Storage


NRC is major participant in two Hawaii
-
based telescopes


CFHT


Canada France Hawaii Telescope


Gemini


Hawaii and Chile


NRC Herzberg will be main repository for telescope data


Initial data volumes in 2005 will be over 7 Terabytes per month


Increasing to 100 Terabytes per month by 2006


Collaborative project involving NRC, CANARIE, the
Australian research network AARnet, National Lambda Rail
in U.S. to swap lightpaths in order to make all the
connections from the telescopes to Canada


Neptune Grid


Joint US
-
Canadian project to build
large undersea fiber network off
west coast of US and Canada


Undersea network will connect
instrumentation devices, robotic
submarines, sensors, under sea
cameras, etc


Distributed computing and data
storage devices on CA*net 4 and
Internet 2 will be used to analyze
and store data


U.S. “Cyberinfrastructure”

Information technology
-
enabled systems, tools and services have
accelerated science and engineering research and education.
Integrated to create a national cyberinfrastructure, these systems,
tools and services are enabling individuals, groups and organizations
to advance science and engineering in ways that revolutionize what
they can do, how they do it, and who can participate.

www.nsf.gov

U.S. “Cyberinfrastructure”

CI research itself includes building new networking protocols,
determining how to discover and access remote resources, providing
programming abstractions of distributed resources to make them
appear local, and understanding how science can take advantage of
the networked world. I want to call out explicitly that a frontier
component of CI is on the "edges," where researchers are extending
the physical infrastructure to sensors and sensor networks wirelessly.

Peter, Arzberger, Chair of the PRAGMA steering committee in GRIDtoday

U.S. “Cyberinfrastructure”

To realize the enormous potential [of CI], new collaborations and
partnerships are needed. Tighter bonds must be forged among the
science and engineering community to ensure the broadest possible
participation among researchers. New collaborations with the private
sector are critical, as they will create the equipment that underpins
tomorrow’s cyberinfrastructure. Finally, new partnerships will be needed
across the globe, to glean insight, ensure openness, interoperability
and a high level of trust within the international community.

Arden L. Bement, Jr., Director, NSF

E.U.: e
-
Infrastructures


The term e
-
Infrastructure is used to indicate the integrated ICT
-
based
Research Infrastructure in Europe [building on] networks,
supercomputers and storage... The initiative is intended to support the
creation of a framework (political, technological and administrative) for
the easy and cost
-
effective shared use of distributed electronic
resources across Europe
-

particularly for grid computing, storage and
networking.

http://www.cordis.lu/ist/rn/ri
-
cnd/einfrastructures.htm

e
-
Infrastructures Roadmap


In 2004 the EC proposed to have a strategic roadmap for Research Infrastructures for next 10
-
20 yrs.


Four roadmaps: Physical Sci/Engineering, Life Sci., Social Sci/Humanities and e
-
Infrastructures


Key components off the current e
-
infrastructures roadmap (version 9) are:




networking infrastructures
:

The network, including lambda
-
networks, is at the heart of everything



middleware and organisation
:
The middleware and virtual organizations connect the distributed resources, data and
storage facilities in a seamless way. Middleware plays the intermediary role to facilitate a deep integration of individual
components. New processes and procedures have to be devised to alter the way organisations work.




resources
:

The European Science Grid should be populated with a number of resources, covering literally everything that
is of interest to science from computers, large storage facilities, telescopes, satellites, special physics equipment,
weather balloons, lasers, spectrometers, visualization means, large sensor networks, large data collections, artificial
intelligence agents and even people.

http://www.e
-
irg.org/roadmap/eIRG
-
roadmap.pdf

E.U.: Adaptive Services Grid


Project under Sixth Framework Programme


“Over the last few years, the service
-
oriented
-
architecture (SoA)
paradigm has gained massive interest from industry and
academia...it will have massive impact on how we build software
systems in the future”, Mathias Weske, ASG Scientific Coordinator


The use of Grid technology in the underlying infrastructure ensures
efficient resource usage...


The ASG consortium has 22 partners from six EU countries and
Australia


www.asg
-
platform.org

SoA


A set of self contained functional components (of higher level
applications) encapsulated as services


Interaction between services through well
-
defined interfaces


Interaction details that are hidden behind the service interface


Enables new modes of user interaction


Fundamentally alters the meaning of an application and the means
by which data can be retrieved, edited, manipulated and re
-
inserted


An application is now the sum of functions (services) addressable
through APIs (interfaces)


“Mashups” are combinations of disparate services through a
common interface


Peter Brantley; University of California, Digital Library

In brief....


SoA is about putting down the paving, the yellow median
stripe, and the shoulders for a fast road. It's about letting
folks drive it without steering for them. Basically, drivers
need minimal guidelines: stay on the road, don't go in the
wrong direction. By providing simple and excellent
infrastructure, you don't have to design the next
generation car
-

the users will do it for you. You just need
to get out of the way.

Peter Brantley; University of California


Kepler


Scientists typically carry out tasks involving the design and execution of a
series of steps, or workflow


Kepler helps organize and automate scientific tasks, a sort of "scientific robot"


Easy for scientists to create both low
-
level 'plumbing workflows' to move data
around, start jobs on remote computers and create high
-
level data analysis
pipelines chaining together standard or custom algorithms


Researcher begins by identifying and accessing initial data sets, and
proceeds through additional steps using software tools such as modeling and
simulation programs, image processing programs and visualization software


Systematic approach to scientific workflows also intended to promote
communication and collaboration through publishing analyses, models, data
transformation programs and derived data sets, giving scientists a way to
track provenance

Paul Tooby, Bertram Ludäscher, UCSD (
http://kepler
-
project.org/

)

Typical Hierarchical System

Sensor

Sensor

Instrument

Instrument

Sensor

Layer 2 switch

Layer 3 switch/router

LAN

Process

Process

Process

Process

Process

LAN

Data Management System

Security

VPN

USER

Instrument Pod

SoA Approach

Sensor/Instrument

WS**

LAN

WS*

Instrument Pod

LAN

WS*

Log Archive Process 1

WS**

Log Archive Process 2

WS**

ONS15454

WS*

CA*net 4

WS*

CANARIE UCLP

WS*

New Web service

Data Management

System

WS**

New development

Lightpath

WS*

WS

Process

WS**

WS HPC

Process

WS**

USER with
Workflow
software

SoA Registry

Australia’s e
-
Research
Coordinating Committee Interim
Report



Advances in ICT make it easier to access, diffuse, communicate and
manipulate large quantities of data, information and knowledge. This
is a worldwide phenomenon ...


The ability of researchers to access data, research facilities,
computational capacity and capability and to work together online,
through advances in ICT, is a significant enabler.


Australian researchers are already increasingly undertaking
research in collaboration with other researchers, in response to the
growing scale and complexity of research ...This is occurring on a
national, international, public and private sector, and inter
-
disciplinary basis.

Context

Australian Report: Issues

1. Human capabilities


success of the strategic e
-
Research framework will be dependent on people


researchers need easy and structured ways of acquiring basic e
-
Research skills.


researchers need day
-
to
-
day support and high level ICT and information management support


2. Access to data


Digital data in all its manifestations is now the core of modern research and knowledge generation


Stakeholders need to adopt best practice in data management, standards and security


3. Need for linkage of e
-
Research resources



researchers need to seamlessly link networks, repositories, instruments and distributed computers



4. Structural and cultural change



The increasing adoption of e
-
Research will require change to organisational structures and cultures


5. Lack of awareness and support



take
-
up of ICTs and infrastructure has not been equally experienced in all disciplines.


6. More collaboration in research



Australian researchers need to increasingly collaborate in national and international research.

Australian Report

Recommendations


1. The Committee’s final report will include an implementation plan

2. As a matter of priority, the Committee will report back to Ministers on
strategies and actions aimed at addressing the skills issues

3. The e
-
Research strategic framework developed will address the need for an
integrated approach in undertaking further research into e
-
Research
enabling technologies and applications

4. The Committee’s final report will specifically address the delivery of e
-
Research outreach, support and research activities on a national basis.


http://www.dest.gov.au/sectors/research_sector/policies_issues_reviews/key_is
sues/ncris/strategic_roadmap.htm






Thank you !

Are there any questions?