CNGrid: Today and Tomorrow

1
CNGrid Today and Tomorrow

• Depei Qian
• depeiq_at_buaa.edu.cn, Beihang University
• July 14, 2007

2
Major Grid researches in China

• China National Grid
• Supported by the National High-tech RD Program
• CROWN
• Supported by the National Science Foundation of
  China
• ChinaGrid
• Supported by the Ministry of Education

3
China National Grid (CNGrid)
4
Overview of CNGrid

• Key project supported by the National High-Tech
  RD Program (the 863 program)
• Earliest and largest Grid project in China
• 4-year effort
• May 2002 to Dec. 2005
• 100 million RMB government funding from the MOST
• More than 2? associated funds from local
  government, application organizations, and
  industry

5
Major tasks of the project

• Developing a test-bed for the new information
  infrastructureChina National Grid (CNGrid)
• Developing a 4TFlops high performance computer
  for the grid
• Developing a set of grid software to support grid
  applications and grid operation and management
• Developing several application grids
• Productive systems
• Demonstration of the future grid applications

6
CNGrid infrastructure

• Major nodes
• CNIC, CAS, Beijing
• Shanghai Supercomputer Center, Shanghai
• Standard nodes
• Tsinghua University, Beijing
• Institute of Applied Physics and computational
  Mathematics, Beijing
• University of Science and Technology of China,
  Hefei
• Xian Jiaotong University, Xian
• NUDT, Changsha
• Hong Kong University, Hongkong
• CNGrid Operation Center
• The CNGrid Operation Center, CNIC, CAS

7
CNGrid nodes locations
8
CNGrid resources

• Computing resource
• 18 TFlops
• Disk storage
• gt200TB
• Application software
• Databases

9
SCCAS

• Equipped with Lenovo DeepComp 6800, 5.324TFlops
• Average utilization 86
• Applications
• 155 registered users
• Supporting more than 120 important projects

10
SSC

• Equipped with Dawning 4000A, 10.2 TFlops
• Utilization gt80
• About 150 users

11
CNGrid operation support center

• Based on the Computer Network Information
  Center of CAS
• Major tasks
• Operational management of CNGrid
• User management
• Security management
• Software maintenance
• Technical support
• User training

12
Grid software development

• Developing a Grid system software
• Supporting CNGrid operation
• Support applications over grid infrastructure

13
CNGrid GOS architecture
14
Grid Software CNGrid GOS

• CNGrid GOS 2.0 deployed on 8 nodes
• Support heterogeneous HPCs Dawning 4000A,
  DeepComp 6800, Galaxy, etc.
• Support multiple platformsUnix, Linux, Windows,
  Apache, WebSphere, etc.

15
Support application Grids
Manufacture Simulation
Bioinformatics
Geo-survey
Weather forecast
16
CNGrid Applications

• Applications from selected areas
• Resource and Environment
• NGG, DFG, SeisGrid
• Research
• SDG, BAGrid, DDG
• Services
• CMAG, TIG
• Manufacturing
• AviGrid, SimGrid

17
National Geological survey Grid NGG

• Undertaken by China National Geological Survey
  Bureau
• Nation-wide productive grid
• Data exchange and sharing by services
• Grid-enabled resource evaluation
• Shorten the evaluation time from 3 years to 3
  months
• Achievements
• North China ground water resources evaluation
• Mineral resources evaluation

18
NGG environment
19
(No Transcript)
20
Digital Forestry Grid

• Develped by a team lead by the Institute of
  Forest Resources Information
• Supporting project of turning the farmland back
  to forests
• potential evaluation
• project planning
• project evaluation
• Providing Forest resource information services

21
Problems facing

• Data owned by local organizations are distributed
  and hard to collect and share
• Data comes from different source
• Hostorical/current
• Remote sensing
• On-site observation
• Need to be Integrated with GIS information

22
(No Transcript)
23
Project planning
24
Project inspection
25
Forest distribution inspection
26
China Meteorological Application Grid

• Developed by China National Meteorological
  Center
• Multiple sites to form the research platform for
  new weather prediction model research
• Provide new weather forecast services (time and
  location-specific)

27
(No Transcript)
28
Computing platform of CMAG
CADP CMA CA Distributed Packages STI Shanghai
Typhoon Research Inst.
29
The load report of grid resources displayed
dynamically by Ganglia within CMAG
30
Real-time weather forecast

• GRAPES meso-scale model runs every day on CMAG
• Real-time forecast products
• http//grid.cma.gov.cn8080/
• Also computing resources and NWP technique
  comparatively less developed regions can utilize
  the grid resources or even run the NWP models to
  satisfy their forecast service requirement
• Yantai
• Qinghai

31
Flowchart of meso-scale GRAPES model on CMAG
32
Snowfall forecast by GRAPES (Dec. 21, 2004)
33
Snowfall forecasting with GRAPES
--last snowfall in Beijing in 2006
34
500 hPa Geopotential height forecast over China
(2005080800UTC)
35
Scientific Data Grid

• Developed by the Computer Network Information
  Center of CAS
• Integration of more than 200 scientific databases
  from more than 40 institutes across the country
  by means of metadata management
• Enable data grid applications
• Virtual astronomical observatory
• Cosmic ray observatory

36
Scientific data resources

• CAS Scientific database
• 45 institutes
• 388 specialty databases
• gt16TB data

37
SDG system platform

• HPC
• DeepCom 6800
• (59??)
• 1TFLOPS
• Mass storage
• 20TB disk
• 50TB tape
• High speed net
• Gbps connection to CSTNet,NSFCNET

38
Software module structure
39
Scientific Data Grid Portal
40
General metadata management tool

• metadata is tree-like and more flexible than
  fix-column tables, difficult to deal with on web
  UI
• use xml files to store interim results

41
Applications virtual astronomical observatory
42
Application cosmic ray data pre-processing system

• Tibet cosmic ray observatory
• High speed network connection from Tibet to
  Beijing (High Energy Physics Institute and SDG
  center)
• High energy database services on SDG

43
Drug discovery grid DDG

• Developed by Shanghai Institute of Materia Medica
  CAS
• Chemical Compound screening for new drug
  discovery using computer simulation
• Lower cost
• Much faster
• Higher accuracy
• Using HPC in P2P mode
• Push and Pull
• Two level job allocation and load ballancing
• New drug for diabetes is under development and
  will enter clinic testing

44
(No Transcript)
45
New Drug Discovery Grid Platform
??? ?
46
DDG Portal
47
Task allocation in DDG
48
Bioinformatics Applied Grid

• Developed by Beijing Genomics Institute, CAS
• developed on the GOS platform
• Composed of data grid, computing grid and
  knowledge grid.
• Various bioinformatics services

49

• Data Grid
• Rice Genome
• Chicken Variations Data
• Silkworm Genome Data
• Computing Grid
• Many Biological Computing Services
• 5 Cities, 6 HPC nodes
• Knowledge Grid
• open genome annotation
• sequencing cooperation

50
Achievements

• Genome data sharing(chicken, silkworm, rice)
• Computing services(Blast /sim4 /BGF /Blat)
• More than 600 computing jobs per day
• Genome sequencing(silkworm, chicken)
• Genome annotation
• Papers in Science and Nature

51
(No Transcript)
52
http//biogrid.genomics.org.cn
53
Aviation Grid AviGrid

• Developed by AVIC II Group of China
• Supporting distributed design, simulation, and
  manufacturing
• CAD software sharing
• Data exchange and sharing
• Computing resource sharing

54
AviGrid
Aviation grid platform composed of three clusters
in different locations
55
AviGrid
Windows
Windows
Automatically locating resource when submitting
jobs
Linux
Windows
AIX
Linux
Linux
Windows
Windows
56
AviGrid
MultiCluster
WAN
Licence pool
GLB
Software sharing
57
AviGrid Topological optimization
After optimization
before optimization
Reduced skin
Added skin
58
The new 863 key project

• Period 2006-2010, launched in Oct. 2006
• 640 million (RMB) from the MOST and at least
  equal amount of matching funding from different
  organizations
• Major efforts
• High productivity Computer development
• CNGrid development
• CNGrid GOS continuous development
• 12-15 Grid applications

59
Objectives

• Capabilities in HPC design and manufacturing
• Innovation on key Grid technologies
• Mature and robust Grid software
• Service oriented Grid infrastructure
• Productive Grid applications

60
High productivity Computer development

• Two phase development
• Two 100TFlops machines in the first phase (By the
  end of 2008)
• One 1000Tflops in the second phase (TBD)
• Installation in CNGrid Major sites
• Promoting HPC and grid applications

61
Grid software development

• Major tasks
• Robust grid system software
• Data grid software
• Grid workflow
• Grid user interface
• Grid security
• Grid management
• Deployment over CNGrid
• Supporting various grid applications

62
Developing Grid service environment

• Upgrading of CNGrid resources
• increased to 12 nodes
• Aggregated computing power of 300TF
• 1PB storage
• Grid monitoring and management
• Strengthen CNGrid operation center
• Management regulation
• CNGrid will become infrastructure for various
  applications

63
Grid applications

• Choose important domains, support key business
• Meteo, Sci Data, Railway, Drug discovery,
  Computational chemistry, astronomy, Chinese
  medicine
• Demonstrate grid technologies
• Achieve productive operations

64
Challenges

• Robust and reliable grid infrastructure
• Mature and robust grid software
• Grid economical and Business model
• Data Grid technologies
• Grid management
• Grid programming model languages
• Grid application development environment
• Grid user interface
• Gridization of traditional applications
• Grid standards

65
CNGrid (2006-2010)

• HPC Systems
• 100 Tflop/s by 2008,
• Pflop/s by 2010?
• Grid Software CNGrid GOS
• Merge with international efforts
• Emphasize matureness
• CNGrid Environment
• Nodes, Centers, mgt
• Applications
• Research
• Resource Environment
• Manufacturing
• Services

66
Cooperation with EU FP6

• EUChinaGrid
• Interoperability between CNGrid GOS and gLite
• IPv6 support
• Application in HEPH, Biology and bioinformatics
• OMII-Europe
• Grid standardization effort
• Interoperbility between main stream grid
  middleware
• EchoGrid
• Grid roadmap
• Coordination between EU and Chinese grid
  researchers
• BRIDGE
• Interoperability between CNGrid GOS and GRIA
• Grid applications Aerospace, Phama, Meteo
• XtreemOS
• Network centric OS

67
Thank you!