Development of an Interdisciplinary fragment of the Russian Grid segment: state of the art

1
Developmentof an Interdisciplinary fragmentof
the Russian Grid segment state of the art
A.Joutchkov, N.Tverdokhlebov, S.Arnautov,
A.Yanovsky, Y.Lyssov, A.Cherny
Telecommunication Center Science
Society Institute for Chemical Physics
RAS Institute for Molecular Biology RAS
2
Telecommunication Centre Science and Society
Moscow Communication Backbone
Ministry of Science
Scientific Society
Association of Scientific Societies of Russia
State Institute of Information Technologies and
Telecommunications
Open Society Institute
Science and Society
3
Our areas of interest

• Telecommunication Grid infrastructure Gigabit
  backbone development and maintenance.
• Biomedical metacomputing Grid-based decisions
  for large-scale computing tasks in biology and
  health care.
• Investigating Grid-related aspects of knowledge
  networks (KN).
• Grid-services providing using specialized portals
  for interested users (scientific organizations
  mainly).

4
RGridin Moscow
Institute for Virusology
Institute for Biomedical Chemistry
SINP MSU Institute for
Nuclear Physics at MSU Science Society There
we
are IOC Institute for Organic
Chemistry EIMB Institute for Molecular
Biology ITEP Institute for Theoretical
and Experimental Physics IITP
International Institute for Theory
of Earthquake Prognosis RSSI
Research Space Science
Institute ?9 Gateway to World
- Prospective
Moscow Medical Academy
National Research Center for Surgery
State Scientific Center for Immunology
5
Backbone of the RGrid
Gatewayto YuMOS
M9
Gateway toCOSMOS
RbNet
RSSI
IITP
Gateway to RUHEP
Science Society320 SI95, 0.28 TB
SINP MSU
EIMB RAS
ITEP
IOC RAS
ICP RAS
GENOMEResources
Gateway to FREENET
GBI RAS
SCC MSU
GigaEthernet 1GB/s
future connections
FastEthernet 100 MB/s
6
GRID InterdisciplinaryProject

• Unifying resources
• Grid Technologies (Globus, )
• Grid Teams (Working groups, Testbeds, )
• Distributed Computation Implementation (LCG,
  BLAST,...)
• Knowledge networks development
  (Genomics,Health Care, Ecology, Chemistry)

7
Science Societyfor Russian HEP Grid
LHC Our
Grid experiments role
activity
ALICE
distributed computations Communication
using RGrid clusters ATLAS backbone,
Grid control CMS
software Testbed 0, Testbed 1,

WP 6 LHCb
8
Science Societyfor Computational Biology
Computational complexity
DNA Length Bacteria 1-3.106 Plants
1-5.108 You and me 4.109
9
BioGrid subsegment
Science Society320 SI95, 0.28 TB
EIMB RAS
IOC RAS
?3524

GBI RAS
ICP RAS IBCP RAS
Farm-central
farm1
Windows Cluster of EIMB
S S Cluster
farm2
farm3
farm4
Institute for Biomedical Chemistry
will be connected in 2002
Virusology Institute
GigaEthernet (1,0 Gb/s)
FastEthernet (100 Mb/s)
10
BLAST experiment on RGridWeb-portal for BLAST
users
Distributed computation schema
11
BLAST experiment on Grid(1st step - without
Globus)
User Query
BLAST-run daemon
Farm-central
Data transfer daemon
GUI
Windows Cluster of EIMB
farm1
Data daemon
BLAST-end daemon
farm2
farm3
Data area
farm4
execution computers
12
BLAST experiment
Search Results
13
BLAST experiment under PBS (2nd step still
without Globus)
User Query
PBS-server
Queue server and Task estimator
Farm-central
GUI www.rgrid.ru
farm1
farm2
farm3
farm4
Windows Cluster of EIMB
execution computers
14
BLAST experiment on Grid(3rd step trying
Globus)
User Query
PBS-server
User autentification and autorization
Farm-central
GRAM
GUI www.rgrid.ru
farm1
farm2
farm3
farm4
Windows Cluster of EIMB
execution computers
15
Knowledge Networks a new level of data
integrationInvestigating Grid-aware aspectsof
KN development
16
Simple-way www-interaction
17
Thin portal(1st step - links to services and
resources)
18
Thick portal (2nd step - searching, collecting
and messaging)
19
Broker ( 3rd element)
20
(No Transcript)
21
Metadata formatsfor Data Collections

• Dublin Core (15 elements)
• Dublin Core with qualificators
• Learning Object Metadata
• MARC (21, Rus)
• Content Standard for Digital Geospatial Metadata
  (FGDC)
• Encoded Archival Description (EAD).

22
What further?

• We are going
• Develop requirements for hardware, software and
  middleware architecture of interdisciplinary
  Grid-node by experiments with Biomedical, HEP and
  Chemical tasks.
• Develop appropriate list of metadata formats to
  construct Knowledge Network prototypes.
• Develop requirements and specifications for
  common and specialized web-interfaces as GUI to
  Grid.