Download Garuda - PRAGMA Grid

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts
no text concepts found
Transcript
Welcome
GARUDA : A National Initiative
of Government of India in Grid
Computing
S. Ramakrishnan
[email protected]
1
Presentation Outline
•
•
•
•
•
Background
Indian Initiatives
C-DAC & PoC Garuda
Grid Partners
Road ahead
2
Presentation Outline
•
•
•
•
•
Background
Indian Initiatives
C-DAC & PoC Garuda
Grid Partners
Discussions
3
Knowledge Paradigm
• High Performance Computing (Petaflop by 2010 and
beyond)
•
Development, acquisition as well as creative use scaling up
momentum
• Petabytes of storages and beyond
•
From Remote Sensing Satellites, MST Radars & Global Scientific
mission forums demanding tremendous storage
• High speed Networks (Terabits per second by 2010 and
beyond)
•
Key enabler to support engineering data and bandwidth intensive
applications
• Grid Computing - supporting distributing computing,
Problem solving environments and collaboration tools
•
Already identified as an most important area and PoC Garuda
under implementation by C-DAC
Knowledge Paradigm (Contd…)
• Data Centre and Web Services
•
Emergence of world class of telecom infrastructure and success of
IT sector augurs well for a host of applications sector from Bioinformatics to E-governance
• Knowledge Management tools
•
Host of with over 300 of Fortune 500 and all top Global ICT MNCs
setting up development centers and increasingly positioning of
research labs in India, as also BPO, KPO centers and VC funding,
GRID Marketing and innovation is happening.
• Secure Cyber Infrastructure
•
Demands for trusted and reliable infrastructure service is
increasing.
• Multilingual Computing
•
With 22 official languages touching over 90% of non-english
speaking people
• Broadband and mobile wireless
•
Fastest growing in India, permeating to villages
Presentation Outline
•
•
•
•
•
06-Sept-2005
Background
Indian Initiatives
C-DAC & PoC Garuda
Grid Partners
Discussions
Presentation to Internet2 & Worldbank
6
Indian Initiatives
Networking, Grid Computing and applications/sectoral domains
•
ERNET
•
IQNET
•
PoC GARUDA
ERNET: Education & Research Network
Started as a collaborative initiative of 8 Premier Institutions
(5 IITs, IISc, NCST, DOE) by Department of Electronics (Now
Department of Information Technology), Government of India
in 1986 with UNDP funding
• Research and development in computer networking
• Built campus LANs, established WAN (terrestrial and
satellite) and first connection from India to Internet
(UUNET) in February ’89.
ERNET: Education & Research Network (Contd…)
Today
•
13 Point of presence (POPs) at premier E&R institutions in the country.
•
STM-1(155Mbps)Ready Fibre-Optic Backbone
•
Satellite Hub in C-band (Bangalore)
•
•
•
•
•
Secure Infrastructure
•
•
•
•
•
•
•
Beaming 3 Transponder of 36 MHz
IP Multicasting
Webcasting
Channel for Distance Learning / Video Broadcasting
Intrusion Detection Server
Firewall
Intruder Alert Manager
Gateway Antivirus Server
Anti Spam Control
Sniffer
Lab dedicated to Network education and training.
Internet connectivity being provided by ERNET
•
User Base
•
•
•
•
•
•
•
172 Universities (250-300)
245 R&D Institutions ( ≥ 500)
52 Engineering colleges (800)
251 Navodya and Govt Schools
274 ICAR Institutions
322 Other educational users/organizations.
Installed Base
•
•
559 TDM/TDMA and 77 SCPC and DAMA VSATs
14 Radio Links and 173 Leased Lines.
INTERNET connectivity under various schemes
AICTE Net
• Connectivity to AICTE recognized colleges and regional
centers. A total of 40 institutions connected. Very Large
potential
UGC Infonet
• MOU signed on 4th April 2003
• 152 universities connected over ERNET backbone.Scalable network
• Multimedia capabilities for video conferencing and
distance learning
ICAR Net
• Network to be implemented in two phases.
• A total of 274 institutions have been connected. Network to
support applications such as VOIP,IP,FAX, Video
conferencing.
NVS Net
• Connectivity provided by VSATS to:
• NVS Head quarters at New Delhi,100 schools have been
ERNET - Geographical Distribution
Univ. of Jammu
Panjab Univ. Chandigarh
Delhi
Univ. of Raj.
Jaipur
(DU)
IIT Guwahati
AMU
IIT Kanpur
CAT Indore
34 Mbps IPLC
Mumbai
VECC
Kolkata
IUCAA Pune
IOP Bhubaneshwar
(TIFR,BARC)
Univ. of
Hyderabad
• Multi-Gigabit pan-European Research Network
• Connecting 32 European Countries and 28 NRENs
IIT Chennai
• Backbone capacity in the range of: 34Mb/s-10Gb/s
AT Austria
DK Denmark
BE Belgium
EE
Estonia
HU Hungary LU Luxembourg
CH Switzerland
ES
Spain
IE
Ireland
LV Latvia
RO Romania
CY Cyprus
FI
Finland
IL
Israel
MT Malta
SE
Sweden
France
IS
Iceland
NL Netherlands
SI
Slovenia
GR Greece
IT
Italy
NO Norway
SK Slovakia
CZ Czech Republic FR
DE Germany
HR Croatia
LT
Lithuania
IISC Banglore
PL
Poland
PT
Portugal
ERNET PoPs
Universities / R&D Institutions
proposed to be connected in
Ist Phase
Additional Links Proposed
ERNET Backbone Links
TR Turkey
UK United Kingdom
IQNET : National QoS Test bed (2005-2007)
• Collaborative effort between C-DAC, ERNET, IITs (Madras,
Bombay, Delhi & Kharagpur)
• QoS Test bed for experimenting with research ideas
• Research activities
•
Measurement Initiative
•
VoIP Initiative
•
Policy based QoS Initiative
• Outcomes expected would include
•
Providing QoS in the Internet
•
Interplay with non QoS networked applications
•
Control and Management of QoS in IP networks
IQNET : Envisaged Connectivity
• Creation of Local test beds
connected to the QoS WAN
Test bed
• QoS WAN Test bed will
overlay on existing ERNET
backbone
• QoS test bed traffic and
regular ERNET traffic logically
separated by running them
over two separate VPNs
ERNET Delhi
LAN
IIT
Delhi
MPLS Cloud
(6PE)
Delhi (P)
Kanpur(6PE)
(6PE)
(6PE)
IIT Mumbai
Kolkata(P)
Mumbai (P)
Kharagpur
IIT
(6PE)
Univ. ofHyd(6PE)
Pune(P)
IIT Chennai (P) IIT Chennai
LAN
Banglore(P)
(6PE)
IISC (6PE)
LAN
CDACBanglore
IISCBanglore
QoS-Net Link
Backbone Link
P
Provider Router
6PE IPv4/IPv6 Enabled
Provider Edge
IQNET : Research Areas in the QoS Testbed
• Development and deployment of technologies and solutions
for distance education
• Experiment to provide application QoS by providing priority,
RSVP and IntServ architecture
• Non Co-operative and Co-operative Measurement and
Characterization
• IP Telephony applications
• Protocol support for Mobile Wireless Endpoints
IQNET : Applications
• Robotic control applications
• Distributed simulation and CAD conferencing
• Applications end point API support for QoS on IPv6
• Telemedicine and Real Time guided clinical investigations
• Deployment of IPTV, H.323 and SIP based telephony and
Content Delivery Network
IQNET : Status
• Measurement Initiative : PingER (Ping End-to-End Reporting)
Collaboration
• Associated with SLAC (Stanford University Linear Accelerator Centre) since
May 06, 2004
• Internet End-to-end Performance Measurement (IEPM) project to monitor endto-end performance of Internet links
• Metrics Measured:
•
•
•
•
response time (rtt(ms))
variability of the response time both short term (time scale of seconds) and longer,
packet loss percentages
and the lack of reachability;
• Content Distribution Initiative: Planet-lab Collaboration
•
•
•
Experimentations explored
MPLS for Linux is a open source effort to create a set of MPLS signaling protocols and
an MPLS forwarding plane for the Linux operating system
Simulation of MPLS using NS-2
IQNET : Status (Contd…)
• Started as a remote node and enhanced to
monitoring node (our’s is one of the 37 monitoring
nodes across the globe)
• Monitoring 59 IPs across the globe
• Sends 100 and 1000 byte icmp packets periodically
• Statistics are stored at our location and reports are
generated
• SLAC maintains the central database
Presentation Outline
•
•
•
•
•
06-Sept-2005
Background
Indian Initiatives
C-DAC & PoC Garuda
Grid Partners
Discussions
Presentation to Internet2 & Worldbank
20
About C-DAC - I
• 10 Locations
• 14 Labs
• 2000 members
About C-DAC – II
R&D areas
•
•
•
•
•
•
High Performance Computing & Grid Computing
• Scientific & Engineering Applications
Multilingual Computing, AAI, Speech Processing &
• Software Technologies, OSS, Multimedia
• ICT for masses
Digital Broadband, Wireless Systems &
• Network Technologies
• e-Security Technologies and Services
Power Electronics, Real-Time Systems &
• Embedded Systems, VLSI/ ANSI Design
Geomatics, Health Informatics, e-Governance &
• Agri Electronics
Education & Training &
• e-Learning Technologies & Services
High Performance Computing
Hardware
•
•
•
•
•
•
Architecture
High-Performance System Design
VLSI Design
System Area Networks (SAN)
Switches
Engineering
System Software & Utilities
•
•
•
•
Compilers
Libraries
Tools
Benchmarking
Application Software
•
•
Scientific & Engineering
Business & Commercial
Advanced Computing, Marketing
and Solutions Group
•
•
•
•
HPC Systems & Technology Consultation
Design and Delivery of HPC facilities and services
End user education and training
Partnering and collaborations
High Performance Computing





Applications Development
High Speed Networks
Reconfigurable Computing
Testing and Certification
HPC Solutions and Training
Garuda – Grid
Computing
2002
Social Computing
with participatory
approach
2007
PARAM Padma
1998
Viable HPC business
computing environment
1994
PARAM 10000
Platform for User community
to interact/ collaborate
1991
PARAM 8000
Technology Denial
MORE
Vision of Grid Computing

Gadgets (Sensors, ……)
Security
Contents :
Data
Engineering,
Data
Management
Device
Electronics
Management
Robustness
People
Generic
Applications
& Tools

Domain
Specific
Applications
& Tools
Distributed, Interconnected, Seemless, Multi-vendor
Computing & Resources Fabric:
Scalability
Hardware, Software, Tools, Data, Instruments
Interoperability
Wireless
Wireline
Indian Grid Computing Initiative
Proof of Concept (PoC) GARUDA phase
• Precursor to the National Grid Computing Initiative (GRID
Garuda)
• Project Duration of 12 months (April 2005 upto March 2006),
starting with Networking Fabric in Collaboration with ERNET,
India
• Major Deliverables
•
•
•
•
•
Technology Development & Research in Grid Computing
Nation-wide high-speed communication fabric
Grid Resources
Deployment of Select applications of National Importance
Grid Strategic User Group
• Implemented by C-DAC
Indian Grid Computing Initiative
Proof of Concept (PoC) GARUDA phase
(Contd..)
• 17 locations till date, 100 Mbps connections with MPLS backbone,
configured for peak load of 2.48 Gbps
• Planned for multidiscipline academic, research and Engineering
applications with some visible demonstratable applications to trigger
progression to main phase (Disaster Management and
Bioinformatics)
• Teraflops of Computing power (including existing 1 Teraflop with CDAC and planned 5 Teraflops early next year), 100s of terabytes
data from various Institutions made available to Grid Parteners
Community
• Intended to migrate smoothly to the main Grid Project from 2006, to
target/address to variety of sectors from basic sciences to major
applications
GRID GARUDA PoC Components
•
•
•
•
Technology Development and Research
Communication Fabric
Computational Resources
Applications
Technology Development & Research
• Technology Deliverables :
•
•
•
•
Architecture & Deployment,
Grid Access Mechanisms,
Application Frameworks,
Problem Solving & Program
Development Environments,
• Grid Middleware and Security,
• Grid Management and Monitoring
• Achievements so far :
• Research Initiatives of Integrated
Development Environments,
• Resource Brokers & Meta Schedulers,
Mobile Agent Framework,
• Semantic Grid Services (with MIT
Chennai)
Communication Fabric
• An ultra-high speed multi services
communication fabric connecting across 17
cities in the country to be deployed jointly by
C-DAC & ERNET.
• Ethernet based High Bandwidth capacity,
Scalable over entire geographic area with
High levels of reliability, fault tolerance and
redundancy.
• Current progress: L2 VPN at 100 Mbps
Connectivity between C-DAC, Pune and
C-DAC, Bangalore.
Module and Cities
Module I
Pune, Bangalore,
Delhi, Kolkata,
Chennai,
Ahmedabad, Mumbai,
Hyderabad
Module II Roorke, Guwahati,
Kharagpur,
Thiruvananthapuram,
Kanpur
Module III Allahabad,
Chandigarh,
Lucknow, Varanasi
Grid Resources
•
Objective is to Provide heterogeneous resources in the Grid including
Compute, Data, Software and Scientific Instruments Deploy Test
facilitates for Grid related research and development activities
•
Deliverables Grid enablement of C-DAC resources at Bangalore and
Pune, Aggregation of Partner Resources Setting up of PoC Test Bed and
Grid Labs at Bangalore, Pune, Hyderabad and Chennai
•
Aggregation of Resources: Prospective partners for grid resources have
been identified & Finalization of specific details under progress
•
Setting up of PoC Test Bed & Grid Labs: Grid Lab equipment has been
received and testing under progress, C-DAC to also set up a grid lab at
SAC, Ahmedabad
Applications of Importance for PoC Garuda
Disaster Management
Application on Garuda
• Objective: Enable applications of
national importance.
• TeraScale Applications
• Weather and Climate modeling
• Seismic Data processing
• Computational Fluid Dynamics
• Structural Mechanics
• Basic Sciences
• Grid-enabled Applications
• Bioinformatics
• Disaster Management
• Data Integration & Sharing
• Earthquake Research
• Cryptanalysis
SAC
Ahmedabad
ASAR flight data
transmission
from
nearby Airport
GRID
Communication
Fabric
High
Speed
Commn
PARAM
Padma
at Bangalore
at Pune
User
Agencies
User
Agencies
Presentation Outline
•
•
•
•
•
06-Sept-2005
Background
Indian Initiatives
C-DAC & PoC Garuda
Grid Partners
Discussions
Presentation to Internet2 & Worldbank
35
Collaborators & Partners : PoC Garuda
C-DAC Centers (10 Locations)
Research Labs
–
–
–
–
–
–
–
Centre for Development of Advanced Computing at :–
–
–
–
–
–
–
–
–
–
Pune,
Bangalore,
Delhi,
Hyderabad,
Mumbai,
Chennai,
Kolkata,
Mohali,
Noida,
Thiruvananthpuram
Academia
–
–
–
–
–
Indian Institute of Science, Bangalore
Madras Institute of Technology, Chennai
University of Pune, Pune
Central University, Hyderabad
Indian Institute of Technology at :–
–
–
–
–
–
–
–
–
–
Kharagpur
Kanpur
Delhi
Mumbai
Chennai &
Guwahati
Guwahati University, Guwahati
Motilal Nehru National Institute of Technology,
Allahabad
Jawaharlal Nehru University, Delhi
Institute of Technology, Banaras Hindu University,
Varanasi
National Chemical Laboratory, Pune
Bhabha Atomic Research Centre, Mumbai
Space Applications Centre, Ahmedabad
Institute for Plasma Research, Ahmedabad
Physical Research Laboratory, Ahmedabad
Saha Institute of Nuclear Physics / Variable Energy
Cyclotron Centre, Chennai
– Regional Cancer Centre, Chennai
– Vikram Sarabhai Space Centre, Chennai
Institutions
– Jawaharlal Nehru Centre for Advanced Scientific
Research, Bangalore
– Indian Institute of Astrophysics, Pune
– National Center for Radio Astrophysics, Pune
– Centre for DNA Fingerprinting and Diagnostics
– Institute of Mathematical Sciences, Chennai
– Institute of Microbial Technology, Chandigarh
– Harish-Chandra Research Institute, Allahabad
– Bhabha Atomic Research Centre, Mumbai
– Central Drug Research Institute, Lucknow
– Sanjay Gandhi Post Graduate Institute of Medical
Sciences, Lucknow
Government Collaborators
– ERNET India
PoC GARUDA Collaborations - In place..
•
SAC, Ahmedabad : Collaboration on Disaster Management and Grid Middleware
•
Indian Institute of Science, Bangalore : Collaboration with Centre for Atmospheric
and Oceanic Sciences (CAOS) Department for Simulations on GRID Garuda
with coupled atmosphere-ocean-land model
•
MIT, Chennai : Collaboration on Grid Middleware Development, Development of
Front End Tools for Grid Services
•
IIT, Mumbai: Collaboration & MoU for porting of CFD solution
•
University of Pune : In application Areas of Quantum Chemistry, Materials
Modeling, Bioinformatics
•
NCL, Pune : Collaboration in the field of Multi-scale Modeling & Simulation,
Large-scale Data Analysis & Mining, HPC & Grid Tools
Where we stand..
• Grid Computing and High Speed Networking: Main Phase of
C-DAC’s Grid Computing project connecting major 200+
Universities, major 300+ R&D/S&T Labs with a backbone of
10+ Gbps, international connectivity of 10 Gbps, 50+ teraflops
of computing power, petaflops of storage, major mission and
sectoral applications
• Planned collaboration in areas of Application, Middleware and
Mission Critical Use for Institutions/Industrial R & D units/labs
to the above, with dependable, consistent, pervasive, secure
and inexpensive access to computational resources
06-Sept-2005
Presentation to Internet2 & Worldbank
38
Mausam GRID
Indian
Coupled GCM-RCM Simulations on
Mausam GRID
Advantages:
i)
General Circulation Models (GCM’s) and
Regional Climate Models (RCM’s) can run on
machines that are physically distributed
ii) Both the models need not be ported to the same
platforms
iii) The models can be owned by different
organizations
Mausam GRID
Application Drivers:
i)
Monsoon Forecasting using GCM
ii) Monsoon Rainfall Downscaling using coupled
atmosphere-ocean system
iii) Extended range monsoon prediction –
multimodel simulations data grid
iv) Coupled regional atmosphere-air quality models
Grid Computing for Bioinformatics
 More than 276 genomes have been sequenced and genome
sequencing of 1220 organisms are at various levels of completion.
 Information retrieved from genome data can prove invaluable for
pharmaceutical industries, for in silico drug target identification and
new drug discovery.
 The enormity of data and complexity of algorithms make the above
tasks computationally demanding, necessitating the effective use of
computational resources beyond those available to researchers at a
single location
 Grid technologies enable sharing of bioinformatics data from different
sites by creating a virtual organization of the data.
Earthquake Research Grid
Main Features • This will connect all the major Earthquake
Engineering centers and some identified high
performance computer centers of India with high
speed network.
Earthquake
Observatory
Data
Server
EE Labs
• This will maintain a database of the digital
earthquake data from different earthquake
observatories and experimental results from
Earthquake laboratories.
Compute
nodes
Compute
nodes
• This will host all the standard software those are
necessary to analyze, process and visualize
earthquake data.
Software
Server
Internet
• Earthquake researchers from remote places can
access this facility through web browsers.
User-
User
• The algorithms developed by the EE researchers
will also be plugged into this facility and make it
available to the other researchers.
User-
-1
3
User2
4
Thank You!
Advanced Computing for Human Advancement
www.cdac.in
06-Sept-2005
Presentation to Internet2 & Worldbank
44