Title: Network Monitoring for Grid Network Operations Center
1Network Monitoring for Grid Network Operations
Center
Seongjin Ahn Sungkyunkwan University sjahn_at_comedu
.skku.ac.kr
2Contents
- Background
- Grid Network
- Grid NOC
- Related study
- Network Measurement
- Grid Monitoring Architecture
- Monitoring Architecture for the Grid NOC
- Goals
- Monitoring Architecture
- Component
- Summary
3Grid
- In the mid1990s to denote a proposed distributed
computing infrastructure for advanced science and
engineering - A communication service that collaborates
dispersed high performance computers, large-scale
databases and modern equipments so that those can
be shared and worked together - Pursing Grid project based Grid network in
several country
4Grid Network
- High performance infrastructure connected
advanced resources for realizing Grid projects - Cooperating with RD networks and primary ISP
networks - High speed network pursuing overseas Grid project
- U.S.A STAR TAP, Abilence, vBNS, NREN
- HGP, NASA IPG, SETI_at_HOME, iVDGL, NEEDSGrid
- European Union GEANT .
- European Data GRID, Euro GRID
- Canada CANet3
- Japan SINET/IMnet
- AP GRID, HEP GRID
- Domestic Network
- KREONet, KOREN, HPCNet, AII-Testbed
5STAR TAP
- The Science, Technology, And Research Transit
Access Point - The premier global exchange point for advanced
international networking, in support of
applications, performance measuring, and
technology evaluations - http//www.startap.net/
6Alliance
- Advanced backbone network that supports
deployment of the new applications being
developed within the Internet2 community - Alliance Grid high speed network to link virtual
environments - http//www.ucaid.edu/abilene/
7NREN
- IPG Baseline System and High Data-rate Testbed
- http//www.nren.nasa.gov/
8GEANT
- GEANT(the pan-European Gigabit Research Network)
taken over TEN-155 - EU DataGrid project
- next generation computing infrastructure
providing intensive computation and analysis of
shared large-scale databases - http//www.dante.net/geant/
9CANet3
- Canada's Research and Education Internet backbone
- Grid Canada
- partnership between CANARIE(Canada's advanced
Internet development organization), the National
Research Council , and C3.ca to grid enable the
nation - http//www.canet3.net/
10Super SINET
- Ultrahigh-speed network intended to develop and
promote Japanese academic - Supercomputer interlocking (GRID) working group
- promoting the construction of
- GRID computing to realize
- computing performance
- http//www.sinet.ad.jp/english/
11TEIN
- Trans Eurasia Information Network
- Contribute to enhancing exchanges and cooperation
between Asia and Europe - Enhance and diversify research exchanges and
cooperation between Asia and Europe - http//www.transeurasia.org/
12KREONet2
- Connect with oversea
- http//www.kreonet2.net/
Korea
USA
Europe
Seoul
APII Testbed KREONet2 (STAR TAP)
Taejon
Pusan
TEIN (Trans-Eurasia Information Network)
chonju
Japan
APII Testbed KREONet2
13Global NOC (1/3)
- Global NOC
- Manage the international network connections from
advanced research and education networks in the
Asia/Pacific, Europe, Russia and South America to
the Science Technology and Research Transit
Access Point (STAR TAP) and the leading US high
performance research and education networks such
as Abilene (the network that supports the
Internet2 project), the NSFs very high
performance Backbone Network System (vBNS) and
the Department of Energys ESNET.
14Global NOC (2/3)
- Principle international connections managed by
the Global NOC - TransPAC
- OC-3 connection between Tokyo and the STAR TAP in
Cicago connecting Japan, Korea, Singapore, and
Australia - Euro-Link
- Multiple high performance connections between The
Netherlans, France, the Nordic countries, Isreal,
and CERN, and the STAR TAP - MIRnet
- Connection between Russia and the STAR TAP
- AmPATH
- Connection to various emerging education and
research networks in Lain America
15Global NOC (3/3)
- Ensuring the reliable provision of the persistent
advanced networking infrastructure that will make
collaboration possible - Functions
- Problem reporting, monitoring, resolution
- Report generation, distribution, archiving
- Web page development, maintenance
- Tool development, maintenance
- http//globalnoc.iu.edu/
16iGOC (iVDGL Grid Operations Center 1/2)
- Global services and centralized monitoring,
management, and support functions - NOC manages a network, providing a single,
dedicated point of contact for iVDGL status,
configuration, and management, and addressing
overall robustness issues - A central monitoring point
- for operational activity
17iGOC (2/2)
- A unified point of support and coordination
- Performance monitoring
- Grid application development
- Grid system incident reporting
- Problem tracking
- To view site specific monitoring information,
provide link to the desired location - http//igoc.iu.edu/
18Grid NOC
- Grid Network Operations Center
- Coordinate global services and centralized
monitoring, management, and technical support to
Grid Network - Allocation of network resources as well as
construction and management of Policy server - Allocation of resources such as CPU and DB
- Collaborative work with international Grid
Network - Research of information resources as well as a
provision education and seminar
19GNOC-kr
- Overall system of GNOC-kr
MIC
Building Operational Structures
Inspecting GRID Making Policies
Resources
GRID Forum
Supercomputer/Cluster/seeMore(CAVE) Science and
Technology Information DB
C
C
GNOC
B
B
- Bandwidth allocation - User
connection control - Network Facility Management - Network
Engineering - Coordination of RD networks for GRID
A
A
Interconnection
Interconnection
HPCNet/KREONET
KOREN
APII-Testbed
- National wide backbone
- (15 regions)
- About 300 RD member institutions
- KR-USA, KR-JP link for RD
- KR-JP, KR- SG links
- KR-USA via JP link
- Backbone over 5 regions
- 32 RD institutions
Domestic RD Networks
A Requiring RD networks for cooperation /
Providing advanced science technology resources
B Requiring advanced science technology
resources / Providing GRID network resources C
User connection control management (Providing
connection channels for local RD networks)
20GNOC-kr
- Connection to overseas GRID networks with
international cooperation
21Collaborative Environments
Environment
meteorology
VAN/USER
CFD
Physics
Drugs
Bio
Middleware
META Computing
Access GRID
Scientific Visualization
GLOBUS
KRISTAL
CAVERN
NOS
NES
NIS
NSS
GRID NOC
GRID Network
SONET/ATM
KOREN
Primary ISPs
KREONet2
STARTAP APII-Testbed
RD NOC
National RD Main Backbone
National RD Access Backbone
International RD Networks
Primary ISPs Testbed
22Network Measurement
- Identifies and characterizes metrics useful to
grid applications and develops standards to
ensure the compatibility of metrics across
measurement systems and applications - Goals
- Describe the metrics of interest to grid
application - Define mappings between available tools and
measurement - Discuss issues in measuring each metric
- Works in collaboration with other standards
groups such as the IETF IPPM WG and the Internet2
End-to-end initiative - IPPMs focus is on defining best-practices
metrics of use of network engineers, whereas the
NMWG focus in cataloging measurement
methodologies as mush as metric
23Network Measurement
24GMA
- Producing a high-level architecture statement of
the components and interfaces needed to promote
interoperability between heterogeneous monitoring
systems on the Grid - Component
- Producer
- Providing event data to consumers
- Publish event availability information in the
directory service - Consumer
- Any process that requests or accepts event data
- Directory Service
- Publish what event data is available and which
producer to contact to get it
25GMA
Consumer
- From http//www-didc.lbl.gov/GGF-PERF/GMA-WG/
Directory Service (LDAP?)
Producer
Plus security!
26Goals
- For GNOC
- Perform overall monitoring with reducing load
- Visualize configuration of overall Grid network
resources - Grasping accurate fault and performance situation
of Grid network - Scalablity as facilitated addition of managed
Grid Network
27Overall architecture
Producer 1
DB
Manager
Directory Service
Producer 2
DB
Producer N
DB
28Directory Service
- Store meta-information of Grid network
- LDAP(Lightweight Directory Access Protocol)
Server - Scalability about new information as hierarchical
management of information - Producer publish location and configuration
information of its domain - Reduce overhead about discovery of whole network
resources
29Producer
- A managing agent having a network domain
- Register managed resource of a network domain in
commonly used directory service for Grid NOC - Collect and analyze fault performance
information - Store management information as putting local DB
in domain
30Producer
- Resource management
- When new resource add, register resource common
directory service of Grid NOC - When resource information delete, delete it from
directory service - Visualization topology
- Visualize overall network constitution with map
- Provide analysis information of fault and
performance on the map - Analysis
- Fault status
- Real-time response to request analysis of fault
and performance - Itemized performance analysis
31Producer
- Reporting
- Keeping of information
- Collection
- Store local DB as collecting information on a
managed domain - Trap daemon
- Store database and real-time forward to
management system - LDAP message handle
- Convert LDAP protocol form registration,
deletion, discovery information - Extract wanted information from response
32Producer
- Message analysis
- Query to resource and compose needed information
from response - Communication
- Set up connection between Directory service and
devices , lines
33Manager
- Operation system that NOCs manager request
information of Grid network - Visualization of configuration regarding overall
network - Connect management system of each network domain,
perform monitoring in detail
34Manager
- Configuration management
- Grasp overall situation of distributed resource
- Effective discovery
- Map Visualization
- Visualize topology of Grid network with map
- Provide analysis information of connection on all
resources - Provide connection with producer
- Directory service communication
- Make users command LDAP protocol
- Connection with directory service
- Make response result wanted information
35Summary
- Effective overall monitoring for Grid NOC
- As producer perform a local management, reduce
load of whole monitoring in Grid NOC - Hierarchically keep location and configuration
information in common directory service ,
guarantee scalablity - Efficiently use resources as managing status of
configuration information about overall network - Ensure safety and efficiency as fault and
performance management of Grid network -