View
217
Download
0
Category
Preview:
Citation preview
7/27/2019 The Grid Introduction
1/19
10/7/201
GRID COMPUTING
Sandeep Kumar PooniaHead Of Dept. CS/IT
B.E., M.Tech., UGC-NETLM-IAENG, LM-I ACSIT,LM-CSTA, LM- AIRCC, LM-SCIEI, AM -UACEE
O UTLINEWhy we are trying to develop acomplex system?Common TermsIntroduction to Grid ComputingMethods of Grid computingGrid MiddlewareGrid Architecture
S a n
d e e p K
um
a r P
o o ni a
Sandeep
Kumar
Poonia
3
W HY WE ARE TRYING TODEVELOP A COMPLEX SYSTEM ?
We can justify the importance of parallel computing fortwo reasons.
Very large application domains, andPhysical limitations of VLSI circuits
Though computers are getting faster and faster, userdemands for solving very large problems is growing ata still faster rate.
Some examples include weather forecasting, simulationof protein folding, computational physics etc. Sandee
pKum
arPoonia
4
P HYSICAL LIMITATIONS OF VLSICIRCUITS
The Pentium III processor uses 180 nano meter(nm) technology, i.e., a circuit element like atransistor can be etched within 180 x 10 -9 m .
Pentium IV processor uses 160nm technology.
Intel has recently trialed processors made by using65nm technology.
Sandee
pKum
arPoonia
5
H OW MANY TRANSISTORS CAN WEPACK ?
Pentium III has about 42 mill ion transistors and
Pentium IV about 55 million transistors.
The number of transistors on a chip isapproximately doubling every 18 months (Moore sLaw).
There are now 100 transistors for every ant onEarth
Sandee
pKum
arPoonia
6
P HYSICAL LIMITATIONS OF VLSICIRCUITS
All semiconductor devices are Si based. It is fairly
safe to assume that a circuit element will take atleast a single Si atom. The covalent bonding in Si has a bond length
approximately 20nm . Hence, we will reach the limit of miniaturization
very soon. The upper bound on the speed of electronic
signals is 3 x 10 8m/sec , the speed of light. Hence, communication between two adjacent
transistors will take approximately 10 -18 sec .
7/27/2019 The Grid Introduction
2/19
7/27/2019 The Grid Introduction
3/19
10/7/201
C LUSTER A RCHITECTURE
S a n
d e e p K
um
a r P
o o ni a
P EER -TO -P EER COMPUTING
Connect to other computersCan access files from any computer on the
network Allows data sharing without going throughcentral serverDecentralized approach also useful for Grid
S a n
d e e p K
um
a r P
o o ni a
P EER TO P EER ARCHITECTURE
S a n d e e p K
um
a r P o
o ni a
W HY G RID C OMPUTING ?
40% Mainframes are idle90% Unix servers are idle95% PC servers are idle0-15% Mainframes are idle in peak-hour70% PC servers are idle in peak-hour
Source: Grid Computing Dr Daron G Green
S a n d e e p K
um
a r P o
o ni a
S a n d e e p K
um
a r P
o o ni a
E LECTRICAL P OWER GRID A NALOGY
Electrical power
gridusers (or electricalappli an ces) get access toelectrici ty th rough wa llsockets with no care orconsideration for where orhow the electricity isactually generated.
The power grid linkstog ether power pl ants of many different kinds
The Grid
users (or cl ien t applications)gain access to computingresources (processors, storage,data, applications, and so on) asneeded with little or noknowledge of where thoseresources are located or whatthe underlying technologies,hardware, operating system,and soon are"the Grid" l ink s tog ethercomputing resources (PCs,workstations, servers, storageelements) and provides themechanism needed to accessthem. Sandeep Kumar Poonia
W HY NEED G RID C OMPUTING ?
Core networking technology now accelerates at amuch faster rate than advances in microprocessorspeedsExploiting under utilized resourcesParallel CPU capacity
Virtual resources and virtual organizations forcollaboration
Access to additional resources
7/27/2019 The Grid Introduction
4/19
10/7/201
Sandeep Kumar Poonia
W HO NEEDS G RID C OMPUTING ?
Not just computer scientistsscientists hit the wall when faced with situations:
The amount of data they need is huge and the data is stored indifferent institutions.The amount of similar calculations the scientist has to do ishuge.
Other areas:GovernmentBusinessEducationIndustrial design
H OW G RID C OMPUTING W ORKS
Super computer,Big mainframe
Idol timeIdol CPU
Idol CPU Idol timeSource: The Evolving Computing Model: Grid Computing Michael Teyssedre
S a n
d e e p K
um
a r P
o o ni a
H OW G RID C OMPUTING W ORKS
Virtual machine Virtual CPU
Idol timeIdol CPU
Idol CPU Idol timeSource: The Evolving Computing Model: Grid Computing Michael Teyssedre
S a n d e e p K
um
a r P o
o ni a
H OW G RID C OMPUTING W ORKS
GridComputing
0% idol0% idol
0% idol 0% idolSource: The Evolving Computing Model: Grid Computing Michael Teyssedre
S a n d e e p K
um
a r P o
o ni a
G RID A RCHITECTURE
Autonomous, globally distributed computers/clusters
S a n d e e p K
um
a r P
o o ni a
W HAT IS A G RID ?
Many definitions exist in the literatureEarly defs: Foster and Kesselman, 1998A computational grid is a hardware and software
infrastructure that provides dependable, consistent,pervasive, and inexpensive access to high-endcomputational facilities
Kleinrock 1969:We will probably see the spread of computer utilities,
which, like present electric and telephone utilities, willservice individual homes and offices across the country.
S a n d e e p K
um
a r P
o o ni a
7/27/2019 The Grid Introduction
5/19
10/7/201
3- POINT CHECKLIST (F OSTER 2002)
1. Coordinates resources not subject tocentralized control
2. Uses standard, open, general purpose protocolsand interfaces3. Deliver nontrivial qualities of service
e.g., response time, throughput, availability,security
S a n
d e e p K
um
a r P
o o ni a
D EFINITION
Grid computing is A distributed computing system
Where a group of computers are connectedTo create and work as one large virtualcomputing power, storage, database, application,and service
S a n
d e e p K
um
a r P
o o ni a
D EFINITIONGrid computing
Allows a group of computers to share the systemsecurely andOptimizes their collective resources to meetrequired workloadsBy using open standards
S a n d e e p K
um
a r P o
o ni a
GRID COMPUTINGGrid computing is a form of distributed computingwhereby a "super and virtual computer" is composed of acluster of networked, loosely coupled computers, acting inconcert to perform very large tasks.
Grid computing (Foster and Kesselman, 1999) is agrowing technology that faci litates the executions of large-scale resource intensive applications ongeographically distributed computing resources.
Faci li tates flexible, secure, coordinated large scaleresource sharing among dynamic collections of individuals, institutions, and resource
Enable communities (virtual organizations ) to sharegeographical ly distributed resources as they pursuecommon goals
S a n d e e p K
um
a r P o
o ni a
A C OMPARISON
SERIAL
Fetch/Store
Compute
PARALLEL
Fetch/Store
Compute/
communicateCooperative game
GRID
Fetch/Store
Discovery of Resources
Interaction with remoteapplication
Authentication / Authorization
Security
Compute/Communicate
Etc
S a n d e e p K
um
a r P
o o ni a
D ISTRIBUTED COMPUTING VS . GRID
Grid is an evolution of distributed computingDynamic
Geographically independentBuilt around standardsInternet backbone
Distributed computing is an older termTypically built around proprietarysoftware and networkTightly couples systems/organization
S a n d e e p K
um
a r P
o o ni a
7/27/2019 The Grid Introduction
6/19
7/27/2019 The Grid Introduction
7/19
10/7/201
G RID T OPOLOGIES
Intragrid Local grid within an organisation
Trust based on personal contracts Extragrid
Resources of a consortium of organisationsconnected through a (Virtual) Private
Network Trust based on Business to Business
contracts Intergrid
Global sharing of resources through theinternet
Trust based on certification
S a n
d e e p K
um
a r P
o o ni a
C OMPUTATIONAL G RID
A computational grid is a hardware and softwareinfras tructure that provides dependable, consistent ,
pervasive, and inexpensive access to high-endcomputational capabilities.
Example : Science Grid (US Department of Energy)
S a n
d e e p K
um
a r P
o o ni a
D ATA G RID A data grid is a grid computing system that deals withdata the controlled sharing and management of large amounts of distributed data .
Data Grid is the storage component of a gridenvironment. Scientific and engineering applicationsrequire access to large amounts of data, and often thisdata is widely distributed. A data grid providesseamless access to the local or remote data required tocomplete compute intensive calculations.
Example :
Biomedical informatics Research Network (BIRN),the Southern California earthquake Center (SCEC).
S a n d e e p K
um
a r P o
o ni a
M ETHODS OF G RIDC OMPUTING
Distributed SupercomputingHigh-Throughput ComputingOn-Demand ComputingData-Intensive ComputingCollaborative ComputingLogistical Networking
S a n d e e p K
um
a r P o
o ni a
D ISTRIBUTED S UPERCOMPUTING
Combining multiple high-capacity resources ona computational grid into a single, virtualdistributed supercomputer.
Tackle problems that cannot be solved on asingle system.Examples: climate modeling, computationalchemistryChallenges include:
Scheduling scarce and expensive resourcesScalability of protocols and algorithmsMaintaining high levels of performance acrossheterogeneous systems
S a n d e e p K
um
a r P
o o ni a
H IGH -T HROUGHPUT COMPUTING
Uses the grid to schedule large numbers of loosely coupled or independent tasks, with thegoal of putting unused processor cycles towork.Schedule large numbers of independent tasksGoal: exploit unused CPU cycles (e.g., fromidle workstations)Unlike distributed computing, tasks looselycoupledExamples: parameter studies, cryptographicproblems
S a n d e e p K
um
a r P
o o ni a
7/27/2019 The Grid Introduction
8/19
7/27/2019 The Grid Introduction
9/19
10/7/201
S a n
d e e p K
um
a r P
o o ni a
TECHNOLOGYTRENDS
Storage, Networks, and Computing power, doubles or,more or less equivalently, halves in price in around 12,9, and 18 months, respectively
S a n
d e e p K
um
a r P
o o ni a
GRID ARCHITECTURE
S a n d e e p K
um
a r P o
o ni a
GRID ARCHITECTURE .
At the lowest level, the fabric , we have the physical
devices or resources that Grid users want to share
and access, including computers, storage systems,
catalogs, networks, and various forms of sensors.
S a n d e e p K
um
a r P o
o ni a
The resource layer contains protocols that exploit communication
and authentication protocols to enable the secure initiation,
monitoring, and control of resource-sharing operations.
Running the same program on different computer systems
depends on resource layer protocols.The Globus Toolkit is a commonly used source of connectivity
and resource protocols and APIs.
S a n d e e p K
um
a r P
o o ni a
The collective layer contains protocols, services, and APIs thatimplement interactions across collections of resources.
Because they combine and exploit components from therelatively narrower resource and connectivity layers, thecomponents of the collective layer can implement a wide varietyof tasks without requiring new resource-layer components.
Examples of collective services includedirectory and brokering services for resource discovery and
allocation;monitoring and diagnostic services;data replication services; andmembership and policy services for keeping track of who in
a community is allowed to access resources.
S a n d e e p K
um
a r P
o o ni a
At the top of any Grid system are the user applications, which are
constructed in terms of, and call on, the components in any otherlayer.For example, a high-energy physics analysis application that needs
to execute several thousands of independent tasks, each taking asinput some set of files containing events, might proceed by
obtaining necessary authentication credentials ;querying an information system and replica catalog to determine
availability of services;submitting requests to appropriate computers, storage systems, and
networks to initiate computations, move data, and so forth (resourceprotocols); and
monitoring the progress of the various computations and datatransfers, notifying the user when all are completed, and detecting andresponding to failure conditions (resource protocols).
7/27/2019 The Grid Introduction
10/19
10/7/201
S a n
d e e p K
um
a r P
o o ni a
AUTHENTICATION, AUTHORIZATION AND POLICY
In Grid environments, the situation is more complex. The distinction
between client and server tends to disappear, because an individual
resource can act as a server one moment (as it receives a request)
and as a client at another (as it issues requests to other resources).
Managing that kind of transaction turns out to have a number of interesting
requirements, such as:
Single sign-on
Mapping to local security mechanisms
Delegation
Community authorization and policy
S a n
d e e p K
um
a r P
o o ni a
AUTHENTICATION, AUTHORIZATION AND POLICY
S a n d e e p K
um
a r P o
o ni a
AUTHENTICATION, AUTHORIZATION AND POLICY
Single sign-on: A single computation may entail access to many resources, but
requiring a user to re-authenticate on each occasion (by, e.g., typing
in a password) is impractical and generally unacceptable.
Instead, a user should be able to authenticate once and then
assign to the computation the right to operate on his or her behalf,
typically for a specified period.
This capability is achieved through the creation of a proxy
credential.
S a n d e e p K
um
a r P o
o ni a
In Figure, the program run by the user (the user proxy)
uses a proxy credential to authenticate at two different
sites.
These services handle requests to create new processes.
S a n d e e p K
um
a r P
o o ni a
AUTHENTICATION, AUTHORIZATION AND POLICY
Mapping to local security mechanisms:Different sites may use different local security solutions, such as
Kerberos and Unix. A Grid security infrastructure needs to map to these local solutions at
each site, so that local operations can proceed with appropriateprivileges.
In Figure, processes execute under a local ID and, at site A, are assigned a
Kerberos ticket, a credential used by the Kerberos authentication system to
keep track of requests.
S a n d e e p K
um
a r P
o o ni a
AUTHENTICATION, AUTHORIZATION AND POLICY
Delegation:
The creation of a proxy credential is a form of delegation, anoperation of fundamental importance in Grid environments.
A computation that spans many resources creates sub-computations(subsidiary computations) that may themselves generate requests toother resources and services, perhaps creating additional sub-computations, and so on.
7/27/2019 The Grid Introduction
11/19
10/7/201
S a n
d e e p K
um
a r P
o o ni a
AUTHENTICATION, AUTHORIZATION AND POLICY
In Figure, the two sub-computations created at sites A and B bothcommunicate with each other and access files at site C. Authentication operations and hence further delegated credentials
are involved at each stage, as resources determine whether to grantrequests and computations determine whether resources aretrustworthy.
The further these delegated credentials are disseminated, thegreater the risk that they will be acquired and misused by anadversary. These delegation operations and the credentials that enablethem must be carefully managed.
S a n
d e e p K
um
a r P
o o ni a
Community authorization and policy: In a large community, the policies that govern who can use which
resources for what purpose cannot be based directly on individual
identity.
It is infeasible for each resource to keep track of community
membership and privileges.
Instead, resources (and users) need to be able to express policies in
terms of other criteria, such as group membership, which can be
identified with a cryptographic credential issued by a trusted third
party.
AUTHENTICATION, AUTHORIZATION AND POLICY
S a n d e e p K
um
a r P o
o ni a
In the scenario depicted in Figure, the file server at site C must know
explicitly whether the user is allowed to access a particular file. A
community authorization system allows this policy decision to be
delegated to a community representative.
G RID M IDDLEWAREGrids are typically managed by grid ware -
a special type of middleware that enable sharing andmanage grid components based on user requirementsand resource attributes (e.g., capacity, performance)Software that connects other software components orapplications to provide the following functions:
Run applications on suitable available resources Brokering, SchedulingProvide uniform, high-level access to resources Semantic interfaces Web Services, Service Oriented Architectures
Address inter-domain issues of security, policy, etc.
Federated IdentitiesProvide application-level statusmonitoring and control
S a n d e e p K
um
a r P o
o ni a
M IDDLEWARES
Globus chicago UnivCondor Wisconsin Univ High throughputcomputingLegion Virginia Univ virtual workspaces-collaborative computingIBP Internet back pane Tennesse Univ logistical networkingNetSolve solving scientific problems inheterogeneous env high throughput & dataintensive
S a n d e e p K
um
a r P
o o ni a
G RID U SERSMany levels of users
Grid developersTool developers
Application developersEnd usersSystem administrators
S a n d e e p K
um
a r P
o o ni a
7/27/2019 The Grid Introduction
12/19
10/7/201
S OME G RID CHALLENGESData movementData replication
Resource managementJob submission
S a n
d e e p K
um
a r P
o o ni a
SOME OF THE M AJOR GRID P ROJECTS
Name URL/Sponsor Focus
EuroGrid, GridInteroperability(GRIP)
eurogrid.orgEuropean Union
Create tech for remote access to super comp resources & simulation codes; inGRIP, integrate with Globus Toolkit
Fusion Collaboratory fusiongrid.orgDOE Off. Science
Create a national computationalcollaboratory for fusion research
Globus Project globus.orgDARPA, DOE,NSF, NASA, Msoft
Research on Grid technologies;development and support of GlobusToolkit; application and deployment
GridLab gridlab.orgEuropean Union
Grid technologies and applications
GridPP gridpp.ac.ukU.K. eScience
Create & apply an operational grid within theU.K. for particle physics research
Grid ResearchIntegration Dev. &Support Center
grids-center.orgNSF
Integration, deployment, support of the NSFMiddleware Infrastructure for research &education
S a n
d e e p K
um
a r P
o o ni a
S a n d e e p K
um
a r P o
o ni a
Grid in India- GARUDA
GARUDA is India's Grid Computinginitiative connecting 17 cities across thecountry.The 45 participating institutes in thisnationwide project include all the IITs andC-DAC centers and other major institutesin India.
S IMULATION TOOLS
GridSim job schedulingSimGrid single client multiserverschedulingBricks schedulingGangSim- Ganglia VOOptoSim Data Grid SimulationsG3S Grid Security services Simulator security services
S a n d e e p K
um
a r P o
o ni a
1 0 / 7 / 2 0 1
3
GT-OGSA Grid Service Infrastructure
OGSI Spec Implementation Security Infrastructure
System-Level Services
Base Services
User-Defined Services
Grid Service Container
Hosting Environment
Web Service Engine
71
S a n d e e p K
um
a r P
o o ni a
1 0 / 7 / 2 0 1
3
THE SPECIFICATION DEFINES HOW E NTITIES CAN C REATE , D ISCOVER AND INTERACT WITH A GRID SERVICE
Servicedata
element
Servicedata
element
Servicedata
element
Service Implementation
GridService(required) other interfaces (optional) Optional:
- Service creation- Notification- Registration- Service Groups
+ application-specific interfaces
Required:- Introspection
(service data)- Explicit destruction- Soft-state lifetime
GT3 Core: OGSI Specification
Includes 0 or more Grid Service Handles (GSHs)Includes 0 or more Grid Service References (GSRs)
Service locator
7 2
Sandee
pKum
arPoonia
http://en.wikipedia.org/wiki/Indiahttp://en.wikipedia.org/wiki/Grid_Computinghttp://en.wikipedia.org/wiki/Indian_Institutes_of_Technologyhttp://en.wikipedia.org/wiki/C-DAChttp://en.wikipedia.org/wiki/C-DAChttp://en.wikipedia.org/wiki/C-DAChttp://en.wikipedia.org/wiki/C-DAChttp://en.wikipedia.org/wiki/Indian_Institutes_of_Technologyhttp://en.wikipedia.org/wiki/Grid_Computinghttp://en.wikipedia.org/wiki/Grid_Computinghttp://en.wikipedia.org/wiki/Grid_Computinghttp://en.wikipedia.org/wiki/India7/27/2019 The Grid Introduction
13/19
10/7/201
1 0 / 7 / 2 0 1
3
Client
A S ERVICE CREATION SCENARIO *
Registry
* The scenarios in this presentation are offered as examples and are not prescriptive
1. From a knownregistry, the clientdiscovers a factory
by querying theService data of theregistry
7 3
Sandee
pKum
arPoonia
1 0 / 7 / 2 0 1
3
Client
Registry
2. The client calls thecreateServiceoperation on thefactory
Factory
1. From a knownregistry, the clientdiscovers a factory
by querying theService data of theregistry
7 4
Sandee
pKum
arPoonia
A S ERVICE CREATION SCENARIO *
1 0 / 7 / 2 0 1
3
Client
Registry
1. From a knownregistry, the clientdiscovers a factory
by querying theService data of theregistry
2. The client calls thecreateServiceoperation on thefactory
Factory
Service
3. The factorycreates aservice
7 5
Sandeep
Kumar
Poonia
A S ERVICE CREATION SCENARIO *
1 0 / 7 / 2 0 1
3
Client
Registry
2. The client calls thecreateServiceoperation on thefactory
Factory
Service
3. The factorycreates aservice
4. The factoryreturns a locator
1. From a knownregistry, the clientdiscovers a factory
by querying theService data of theregistry
7 6
Sandeep
Kumar
Poonia
A S ERVICE CREATION SCENARIO *
1 0 / 7 / 2 0 1
3
Client
Registry
2. The client calls thecreateServiceoperation on thefactory
Factory
Service
3. The factorycreates aservice
4. The factoryreturns a locator
5. The client and service interact
1. From a knownregistry, the clientdiscovers a factory
by querying theService data of theregistry
7 7
Sandee
pKum
arPoonia
A S ERVICE CREATION SCENARIO *
1 0 / 7 / 2 0 1
3
NotificationSink
A N OTIFICATION SCENARIO
1. NotificationSink calls thesubscribe operation on
NotificationSource
NotificationSource 7
8
Sandee
pKum
arPoonia
7/27/2019 The Grid Introduction
14/19
7/27/2019 The Grid Introduction
15/19
10/7/201
1 0 / 7 / 2 0 1
3
GT-OGSA Grid Service Infrastructure
OGSI Spec Implementation Security Infrastructure
System-Level Services
Base Services
User-Defined Services
Grid Service Container
Hosting Environment
Web Service Engine
85
S a n
d e e p K
um
a r P
o o ni a
1 0 / 7 / 2 0 1
3
GT3 C ORE : S YSTEM LEVEL SERVICES
General-purpose services that facilitate the use of GridServices in production environmentsThe 3.0 distribution includes the following System-Levelservices:
An Administration Service A Logging Service A Management Service
8 6
Sandee
pKum
arPoonia
1 0 / 7 / 2 0 1
3
GT-OGSA Grid Service Infrastructure
OGSI Spec Implementation Security Infrastructure
System-Level Services
Base Services
User-Defined Services
Grid Service Container
Hosting Environment
Web Service Engine87
S a n d e e p K
um
a r P o
o ni a
1 0 / 7 / 2 0 1
3
GT3 C ORE : G RID S ERVICE C ONTAINER
Interface Layer
Transport Layer
Implementation Layer
Layers in the Web Services Model
OGSI Spec is here
Transport/BindingLayer (GT3 supportsSOAP over HTTP)
Container is here
8 8
Sandeep
Kumar
Poonia
1 0 / 7 / 2 0 1
3
GT-OGSA Grid Service Infrastructure
OGSI Spec Implementation Security Infrastructure
System-Level Services
Base Services
User-Defined Services
Grid Service Container
Hosting Environment
Web Service Engine
89
S a n d e e p K
um
a r P
o o ni a
1 0 / 7 / 2 0 1
3
GT3 C ORE : H OSTING E NVIRONMENT
GT3 currently offers support for four JavaHosting Environments:
EmbeddedStandaloneServletEJB
9 0
Sandee
pKum
arPoonia
7/27/2019 The Grid Introduction
16/19
10/7/201
1 0 / 7 / 2 0 1
3
GT3 C ORE : V IRTUAL H OSTING E NVIRONMENTF RAMEWORK
Virtual Hosting allows grid services to bedistributed across several remote containers
Useful in implementing solutions for problemscommon to distributed computing
Load balancingUser account sandboxing
9 1
Sandee
pKum
arPoonia
1 0 / 7 / 2 0 1
3
A S ERVICE CREATION S CENARIOILLUSTRATING REDIRECTION IN V IRTUAL H OSTING
Client
Registry
Router
HE Starter
1. Froma knownregistry,the clientretrievesa factorylocator
9 2
Sandee
pKum
arPoonia
1 0 / 7 / 2 0 1
3
A S ERVICE CREATION SCENARIOILLUSTRATING R EDIRECTION IN V IRTUAL H OSTING
Client
Registry
Router 1. Froma knownregistry,the clientretrievesa factorylocator
HE Starter 2. The routerintercepts thecreateServicecall on thefactory
9 3
Sandeep
Kumar
Poonia
1 0 / 7 / 2 0 1
3
A S ERVICE CREATION S CENARIOILLUSTRATING REDIRECTION IN V IRTUAL H OSTING
Client
Registry
Router 1. Froma knownregistry,the clientretrievesa factorylocator
2. The routerintercepts thecreateServicecall on thefactory
HE Starter
3. The router passes the createServicerequest to the Host Env Starter
9 4
Sandeep
Kumar
Poonia
1 0 / 7 / 2 0 1
3
A S ERVICE CREATION SCENARIOILLUSTRATING R EDIRECTION IN V IRTUAL H OSTING
Client
Registry
Router
Service
1. Froma knownregistry,the clientretrievesa factorylocator
2. The routerintercepts thecreateServicecall on thefactory
HE Starter
3. The router passes the createServicerequest to the Host Env Starter
4.The HEStartercreatesa newHost Envas wellas theservice
9 5
Sandee
pKum
arPoonia
1 0 / 7 / 2 0 1
3
A S ERVICE CREATION S CENARIOILLUSTRATING REDIRECTION IN V IRTUAL H OSTING
Client
Registry
Router
Service
1. Froma knownregistry,the clientretrievesa factorylocator
2. The routerintercepts thecreateServicecall on thefactory
HE Starter
3. The router passes the createServicerequest to the Host Env Starter
4.The HEStartercreatesa newHost Envas wellas theservice
5. The router returnsa service locator
9 6
Sandee
pKum
arPoonia
7/27/2019 The Grid Introduction
17/19
7/27/2019 The Grid Introduction
18/19
10/7/201
1 0 / 7 / 2 0 1
3
Client
GRAM J OB SUBMISSION SCENARIO
IndexService
1. From an indexservice, the clientchooses anMMJFS
2. The client calls thecreateServiceoperation on thefactory,supplyingRSL
MMJFS
MJS
3. The factorycreates aManaged JobService
1 0 3
Sandee
pKum
arPoonia
1 0 / 7 / 2 0 1
3
Client
GRAM J OB SUBMISSION SCENARIO
IndexService
1. From an indexservice, the clientchooses anMMJFS
2. The client calls thecreateServiceoperation on thefactory,supplyingRSL
MMJFS
MJS
3. The factorycreates aManaged JobService
4. The factoryreturns a locator
1 0 4
Sandee
pKum
arPoonia
1 0 / 7 / 2 0 1
3
Client
GRAM J OB SUBMISSION SCENARIO
IndexService
1. From an indexservice, the clientchooses anMMJFS
2. The client calls thecreateServiceoperation on thefactory,supplyingRSL
MMJFS
MJS
3. The factorycreates aManaged JobService
4. The factoryreturns a locator
5. The client subscribes tothe MJS status SDE andretrieves output
1 0 5
Sandeep
Kumar
Poonia
1 0 / 7 / 2 0 1
3
GT3 B ASE : I NFORMATION SERVICES
Index Service as Caching AggregatorCaches service data from other grid services
Index Service as Provider FrameworkServes as a host for service data providers that liveoutside of a grid service to publish data
1 0 6
Sandeep
Kumar
Poonia
1 0 / 7 / 2 0 1
3
GT3 B ASE : RELIABLE F ILE TRANSFER
Reliably performs a third party transfer between two GridFTPserversOGSI-compliant service exposing GridFTP control channelfunctionalityRecoverable Grid Service
Automatically restarts interrupted transfers from the last checkpoint
Progress and Restart Monitoring
GridFTPServer 1
GridFTPServer 2
RFT
JDBC
1 0 7
Sandee
pKum
arPoonia
1 0 / 7 / 2 0 1
3
GT-OGSA Grid Service Infrastructure
OGSI Spec Implementation Security Infrastructure
System-Level Services
Base Services
User-Defined Services
Grid Service Container
Hosting Environment
Web Service Engine
108
S a n d e e p K
um
a r P
o o ni a
7/27/2019 The Grid Introduction
19/19
10/7/201
1 0 / 7 / 2 0 1
3
GT3 U SER -D EFINED S ERVICES
GT3 can be viewed as a Grid Service DevelopmentKit that includes:
Primitives designed to ease the task of building OGSI-Compliant ServicesPrimitives for provisioning securityBase services that provide an infrastructure with whichto build higher-level services
1 0 9
Sandee
pKum
arPoonia
1 0 / 7 / 2 0 1
3
GT3 U SER -D EFINED S ERVICES (CONT .)
ANT
User source files
GT3 Build Files
User Build File
GridServiceexecutablefiles
(Diagram inspired byBorja Sotomayorsexcellent tutorial on GT3)
1 1 0
Sandee
pKum
arPoonia
1 0 / 7 / 2 0 1
3
F UTURE D IRECTIONS OF GT
Standardization of container modelDevelopment of lightweight container/api
Adding rich support for queriesFurther refinements of Base Service designsPushing on standardizing at a higher level than OGSI
1 1 1
Sandeep
Kumar
Poonia
Recommended