Grid computing

Grid Computing
Shashwat Shriparv
dwivedishashwat@gmail.com
InfinitySoft

Agenda
► Introduction
► Supercomputer
Example: PARAM padma
► Grid Computing
 Definition, Evolution not revolution, Why Grid Computing Now? …
► Design considerations and variations
 Berkeley Open Infrastructure for Network Computing (BOINC)
► How to implement the concept of grid computing
 Grid computing middleware ,Installation configuration and run.
► Service-Oriented Grid Architecture and Gridbus Solutions
 Utility Networks and Grid Computing, Grid Challenges.
► Some Grid Initiatives

Supercomputer
► A time dependent term which refers to the class
of most powerful computer systems world-wide
at the time of reference.
► An extremely fast computer that can perform
hundreds of millions of instructions per second..
► Examples :PARAM Padma (C-DAC), Columbia
Supercomputer (NASA)

PARAM Padma
►PARAM Padma is C-DAC’s next generation
high performance scalable computing
cluster,currently with a peak computing
power of One Teraflop.
►It has 248 processors (IBM POWER4 @
1GHz ) , 62 nos. of IBM p630 Nodes and
496 GB Main Memory.

specifications
1: No. of Processors 248 (Power 4@1GHz)
2: Internal Storage 4.5 TeraBytes
3: Operating System AIX / LINUX
4: Networks Primary : PARAMNet-II @ 2.5
Gbps Full Duplex
Backup : Gigabit thernet @ 1
Gbps Full Duplex
5: Total Weight 9 Tons
6: Cabling (Networks) 12Kms (CAT6 ) & 2.5Kms(Fibre
Optic)
7: Total Machine Floor Area 168 Sq.metre (1800sq.ft.)

Agenda
► Introduction
Supercomputer
► Grid Computing
 Definition, Evolution not revolution, Why Grid Computing
Now? …

Grid Computing
►Using distributed computers and
resources collectively.
►Usually associated with geographically
distributed computers and resources on a
high speed network.
►Often about teams sharing resources.

► The primary advantage of Grid computing is that
each node can be purchased as commodity
hardware, which when combined can produce
similar computing resources to a many-CPU
supercomputer, but at lower cost.
► The primary performance disadvantage is that the
various CPUs and local storage areas do not have
high-speed connections. This arrangement is thus
well-suited to applications where multiple parallel
computations can take place independently,
without the need to communicate intermediate
results between CPUs.

Evolution, not revolution
► grid can be seen as the latest and most complete evolution
of more familiar developments such as distributed
computing, the Web, peer-to-peer computing and
virtualization technologies.
► Like the Web, grid computing keeps complexity hidden:
multiple users enjoy a single, unified experience.
► Unlike the Web, which mainly enables communication,
grid computing enables full collaboration toward common
business goals.

► Like clusters and distributed computing, grids bring
computing resources together.
► Unlike clusters and distributed computing, which
need physical proximity and operating homogeneity, grids
can be geographically distributed and heterogeneous.
► Like virtualization technologies, grid computing
enables the virtualization of IT resources.
► Unlike virtualization technologies, which virtualize a
single system, grid computing enables the virtualization of
vast and disparate IT resources

► Supercomputer: It can be costly and difficult to
write programs so that they can be run in the
environment of a supercomputer, which may have
a custom operating system.
► Grid computing: If a problem can be adequately
parallelized, a "thin" layer of "grid" infrastructure
can cause conventional, standalone programs to
run on multiple machines (but each given a
different part of the same problem)

Local
Cluster
Inter Planet
Grid
2100
2100 2100 2100 2100
2100 2100 2100 2100
Personal Device SMPs or
SuperComputers
Global
Grid
Enterprise
Cluster/Grid
Scalable Computing
P
E
R
F
O
R
M
A
N
C
E
+
Q
o
S

Agenda
► Introduction
Supercomputer
► Grid Computing
 Berkeley Open Infrastructure for Network Computing
(BOINC)

Design considerations and
variations
► One feature of distributed grids is that they can
be formed from computing resources belonging
to multiple individuals or organizations, so it is
easy to assemble volunteer computing networks.
► One disadvantage of this feature is that the
computers which are actually performing the
calculations might not be entirely trustworthy.
This often involves assigning work randomly to
different nodes and checking that at least two
different nodes report the same answer for a
given work unit.

►Due to the lack of central control over the
hardware, there is no way to guarantee that
nodes will not drop out of the network at
random times. These variations can be
accommodated by assigning large work
units (thus reducing the need for continuous
network connectivity) and reassigning work
units when a given node fails to report its
results as expected.

►Various middleware projects have created
generic infrastructure, to allow various
scientific and commercial projects to utilize
a particular associated grid, or for the
purpose of setting up new grids. BOINC is
a common one for academic projects
seeking public volunteers.

Berkeley Open Infrastructure
for Network Computing
► The Berkeley Open Infrastructure for
Network Computing (BOINC) is a non-
commercial middleware system for grid
computing.
► The intent of BOINC is to make it possible for
researchers to tap into the enormous processing
power of personal computers around the world.
► BOINC has over 4,30,000 active computers (hosts)
worldwide processing on average 663 TFLOPS as
of September 8, 2007.

►Design and structure of BOINC
BOINC client-server technology.
►BOINC Credit System
Validating results before granting credit.
►Projects using BOINC Framework
1.Cell Computing — biomedical research.
2.BBC Climate Change Experiment —
(part of Climateprediction.net)

Agenda
► Introduction
Supercomputer
► Grid Computing

►Grid computing middleware
S/W which is used to manage the
internal working of a grid computing system.
Eg:
1:Alchemi: A .NET-based Enterprise Grid
System and Framework.

How Alchemi Works
There are four types of distributed
components (nodes) involved in the
construction of Alchemi grids and execution
of grid applications: Manager, Executor,
User & Cross-Platform Manager.

►A grid is created by installing Executors on
each machine that is to be part of the grid
and linking them to a central Manager
component.
Executors
Dedicated Nondedicated

► dedicated :-
meaning the Manager initiates thread execution directly.
► non-dedicated:-
meaning that thread execution is initiated by the
Executor.
works through firewalls and NAT servers since there is
only one-way communication between the Executor and
Manager.
► Dedicated Executors are more suited to an intranet
environment
► and non-dedicated Executors are more suited to the
Internet environment.

► Users can develop, execute and monitor
grid applications using the .NET API and
tools which are part of the Alchemi SDK.
Alchemi offers a powerful grid thread
programming model.
►An optional component (not shown in fig:)
is the Cross Platform Manager web service
which offers interoperability with custom
non-.NET grid middleware.

Installation, Configuration and
Operation

► Common Requirements
MANAGER
The Manager should be installed on a stable and
reasonably capable machine. The Manager requires SQL
Server 2000 or MSDE 2000.
Microsoft .NET Framework 1.1

►The Alchemi Manager can be installed in
two modes
1: As a normal Windows desktop application
2: As a windows service. (supported only on
Windows NT/2000/XP/2003)

► Role-Based Security
Every program connecting to the Manager must supply a
valid username and password. Three default accounts are
created during installation: executor (password: executor),
user (password: user) and admin (password: admin)
belonging to the 'Executors', 'Users' and 'Administrators'
groups respectively.
Users are administered via the 'Users' tab of the Alchemi
Console (located in the Alchemi SDK)Only.

Executor
The Alchemi Executor can be installed in two
modes
1: As a normal Windows desktop
application
2: As a windows service. (supported only
on Windows NT/2000/XP/2003)
Installation

Configuration & Operation
You need to configure 2 aspects of the Executor:
• The host and port of the Manager to connect to.
• Dedicated / non-dedicated execution. A non-dedicated
Executor executes grid threads on a voluntary basis (it
requests threads to execute from the Manager), while a
dedicated Executor is always executing grid threads (it is
directly provided grid threads to execute by the Manager).
•A non-dedicated Executor works behind firewalls and
NAT servers.
•Click the "Connect" button to connect the Executor to the
Manager.

If the Executor is configured for non-dedicated
execution, we can start executing by clicking the "Start
Executing" button in the "Manage Execution" tab.

► Alchemi Console
The Console (Alchemi.Console.exe) is a grid
administration and monitoring tool.

Agenda
► Introduction
Supercomputer
► Grid Computing
 Utility Networks and Grid Computing, Grid Challenges

Essential Utilities and
Delivery Networks
(1) Water
Water
Distribution
Network
(4) Telephone
Telecom
Networks
(3) Gas
(2) Electricity
Power Grid

Computing Grid: Delivering IT services
as the 5th utility (Power Grid inspiration)
eScience
eBusiness
eGovernment
eHealth
Multilingual
eEducation
…

Classes of Grid Services / Types of Grids
► Computational Services – CPU cycles
 Pooling computing power: SETI@Home,
TeraGrid, AusGrid, ChinaGrid, IndiaGrid, UK
Grid,…
► Data Services
 Collaborative data sharing generated by
instruments, sensors, persons: LHC Grid,
Napster
► Application Services
 Access to remote software/libraries and license
management—NetSolve
► Interaction Services
 eLearning, Virtual Tables, Group Communication
(Access Grid), Gaming
► Knowledge Services
 The way knowledge is acquired, processed and
managed—data mining.
► Utility Computing Services
 Towards a market-based Grid computing:
Leasing and delivering Grid services as ICT
utilities.
Computational Grid
Data Grid
ASP Grid
Interaction Grid
Knowledge Grid
Utility Grid
infrastructure
Users

Grid Challenges
Security
Resource Allocation
& Scheduling
Data locality
Network Management
System Management
Resource Discovery
Uniform Access
Computational Economy
Application Construction

What do Grid players want?
► Grid Consumers
 Execute jobs for solving varying problem size and
complexity
 Benefit by utilizing distributed resources wisely
 Tradeoff timeframe and cost
►Strategy: minimise expenses
► Grid Providers
 Contribute resources for executing consumer jobs
 Benefit by maximizing resource utilisation
 Tradeoff local requirements & market opportunity
►Strategy: maximise return on investment

Principle 1: Service Oriented
Architecture (SOA)
► A SOA is a contractual architecture for offering and consuming software as
services.
► There are four entities that make up an SOA
 service provider,
 service registry, and
 service consumer (also known as service requestor).
► The functions or tasks that the service provider offers, along with other
functional and technical information required for consumption, are defined
in
 the service definition or contract.
provider
registry
consumer
contract

Principle 2: Market-Oriented (Grid) Computing-
(a) Sustained Resourced Sharing and (b) Effective
Management of Shared Resources
Grid Economy

On Demand Assembly of Services: Putting
Them All Together
ASP Catalogue
Grid Info Service
Grid Market Directory
GSP
(Accounting Service)
Gridbus
GridBank
GSP
(e.g., UofM)
PE
GSP
(e.g., VPAC)
PE
GSP
(e.g., IBM)
CPU
or
PE
Grid Service (GS)
(Globus)
Alchemi
GS
GTS
Cluster Scheduler
8
Grid
Resource Broker
2
Visual Application Composer
Application Code
Explore
data1
36
4
5
9 7
10
11
Bill
12
Data Catalogue

Active Sheet:
Spreadsheet Processing on Grid using Nimrod-
G
Nimrod
Proxy
Nimrod-G
World-Wide Grid

What are Grid benefits?
►Resource sharing across multiple
administrative boundaries
 Effective utilisation of the (existing)
resources
 Dynamic provisioning
►Application Acceleration
►Scalability
►Virtualisation:
 applications, services, resources,…

Some Grid Initiatives Worldwide
► Australia
 Nimrod-G
 Gridbus
 DISCWorld
 GrangeNet.
 APACGrid
 ARC eResearch
► Brazil
 OurGrid, EasyGrid
 LNCC-Grid + many others
► China
 ChinaGrid – Education
 CNGrid - application
► Europe
 UK eScience
 EU Grids..
 and many more...
► India
 Garuda
 Japan
 NAGERI
► Korea...
N*Grid
► Singapore
NGP
► USA
 Globus
 GridSec
 AccessGrid
 TeraGrid
 Cyberinfrasture
 and many more...
► Industry Initiatives
 IBM On Demand Computing
 HP Adaptive Computing
 Sun N1
 Microsoft - .NET
 Oracle 10g
 Infosys – Enterprise Grid
 Satyam – Business Grid
 StorageTek –Grid..
 and many more
► Public Forums
 Global Grid Forum
 Australian Grid Forum
 Conferences:
► CCGrid
► Grid
► HPDC
► E-Sciencehttp://www.gridcomputing.com
1.3 billion – 3 yrs
1 billion – 5 yrs
450million – 5 yrs
1.3 billion (Rs)
27 million
2? billion

► The Department of Information Technology (DIT),
Government of India has funded the Centre for
Development of Advanced Computing (C-DAC) to
deploy the nation-wide computational grid 'GARUDA'
which will connect 17 cities across the country in its
Proof of Concept (PoC) phase with an aim to bring
"Grid" networked computing to research labs and
industry. A total of 45 institutions have been
connected. GARUDA will accelerate India s drive to
turn its substantial research investment into tangible
economic benefits.

►The GARUDA High-Speed network is a
Layer 2/3 MPLS Virtual Private Network
(VPN) connecting select 45 institutions
across 17 cities at 10/100 Mbps with
Stringent Service Level Agreements with
the service provider.

Shashwat Shriparv
dwivedishashwat@gmail.com
InfinitySoft

Grid computing

More Related Content

What's hot (20)

Viewers also liked (12)

Similar to Grid computing (20)

More from Shashwat Shriparv (20)

Grid computing