GRID COMPUTING

1.1Introduction:-

IAN FOSTER
 When the network is as fast as the computer's internal links, the machine
disintegrates across the net into a set of special purpose appliances.

-- Gilder Technology Report, June 2000.

“United we stand, divided we fall”, is the key idea behind grid computing. Grid is a
type of parallel and distributed system that enables the sharing, selection, and aggregation of
geographically distributed "autonomous" resources dynamically at runtime depending on
their availability, capability, performance, cost, and users' quality-of-service requirements.
Grid computing is a critical shift in thinking about how to maximize the
value of computing resources. It allows us to unite pools of servers, storage systems and
networks into a single large system so we can deliver the power of multiple-systems resources to
a single user point for a specific purpose. To a user, data file, or an application, the system
appears to be a single, enormous virtual computing system.
Grid computing is the next logical step in distributed networking. Just as the
Internet allows users to share ideas and files as the seeds of projects, grid computing lets us share
the resources of disparate computer systems. The major purpose of a grid, is to virtualize
resources to solve problems. So, rather than using a network of computers simply to
communicate and transfer data, grid computing taps the unused processor cycles or numerous i.e.
thousands of computers.

1.2 Definition:- A means of network computing that harnesses the unused processing cycles
of numerous computers, to solve intensive problems that are often too large for a single computer
to handle, such as in life sciences or climate modeling.

1.3 Origin:-
Its roots lies in early distributed computing projects that date back to the 1980s As a
result, grid has become a centerpiece of the "utility computing" marketing drive taken up by
nearly every vendor.

1.4Phases [history]
a) First generation:-FAFNER(Factoring via Network-Enabled Recursion) and I-WAY are the
two first generation grid computing methods. FAFNER is Based on public key encryption
algorithm, was used factorize extremely large numbers

b) Second generation core technologies

GLOBUS:
It is based on globus toolkit resource allocation manager, an extended version of FTP protocol,
which enables data access, partial file access management at high speed.
c) Third generation systems
These system follows properties of autonomy such as
1. Detail knowledge of its component and status
2. Must configure reconfigure itself dynamically
3. able to recover from malfunction
4. Protect itself against attack
The third generation grid computing involves use of different techniques such as
XML, java, J2EE etc.
Different Types:-Distributed object system it is based on Common Request Broker Architecture
(CORBA) . We can use java with CORBA up to certain extent as it provides distributed object
through RMI. Java has drawbacks in terms of computation speed.
Jini and RMI:-Jini is designed to provide a software infrastructure that can form distributed
computing environment .in this application is usually written in java and communication using
RMI. Key concepts of jini are Lookup, Discovery, Leasing , Remote events and Transaction .
1.5 Need :-The best candidates for grid are applications that run the same or similar
computations on thousands or millions of pieces of data, with no single calculation dependent on
those that came before.

2. Grid:-

GRIDS ARE USUALLY HETEROGENEOUS NETWORKS. GRID NODES,

GENERALLY INDIVIDUAL COMPUTERS, CONSIST OF DIFFERENT
HARDWARE AND USE A VARIETY OF OPERATING SYSTEMS AND
NETWORKING TO CONNECTING THEM VARY IN BANDWIDTH. THESE
RESOURCES ARE USED AMONG THE VARIOUS PROJECTS. THIS FORMS THE
SYSTEM AS THE AGGREGATION OF RESOURCES FOR A PARTICULAR TASK
I.E. VIRTUAL ORGANIZATION.

Simple Grid Diagram

3. Types of Grid:-
3.1Computational Grid:- It is focused on settings aside resources specifically for
computing power .In this type of grid most of machines are high performance servers.
3.2 Scavenging Grid:-It is most commonly used with large numbers of desktop machine.
Machines are scavenged for available CPU cycles and other resources. Owners of desktop
machines are usually given control over when their resources are available to grid.
3.3 Data Grid:-It is responsible for housing and providing access to data across multiple
organizations. Users are not concerned with where this data is located as long as they access to
the data .

4 HOW IT WORKS?
Grid computing uses networked clusters of CPUs connected over the Internet, a
company intranet or a corporate WAN. The resulting network of CPUs acts as a foundation for a
set of grid-enabling software tools. These tools let the grid accept a large computing job and
break it down into tens, hundreds or thousands of independent tasks. The tools then search the
grid for available resources; assign tasks to processors, aggregate the work and spit out one final
result.

A grid user installs the provided grid software (for using the grid as well as
donating to the grid) on his machine and gets connected with Internet. The user establishes his
identity with a certificate authority. This software may be automatically reconfigured by the
grid management system to know the communication address of the management nodes in
the grid and user or machine identification information.

To use the grid, most grid systems require the user to log on to a system using
a user ID that is enrolled in the grid. Once logged on, the user can query the grid and submit
jobs. Grid systems usually provide command line tools as well as graphical user interfaces
(GUIs) for queries. Command line tools are especially useful when the user wants to write a
script.

Job submission usually consists of three parts, even if there is only one command
required.

 Some input data and possibly the executable program or execution script file are sent to
the machine to execute the job. Sending the input is called “staging the input data.”

 The job is executed on the grid machine.

 The results of the job are sent back to the submitter. When there are a large number of
sub jobs, the work required to collect the results and produce the final result is usually
accomplished by a single program

5. a QoS (Quality of Service) guided scheduling algorithm

Scheduler is the main part of grid computing. In the three main phases of scheduler, first is resource
discovery, second phase involves gathering information about resources and choosing the best
match for application requirement. In third phase the job is executed. The scheduling algorithms
are mainly divided into two categories: online mode and batch mode. In Minimum Completion
Time, grid system assigns the task to the machine, that will have earliest completion time, and in
Minimum Execution Time, it assigns the task to the machine that performs task, in least execution
time.

Min-min algorithm

While there are tasks to schedule

For all tasks to schedule
For all hosts
Compute the expected completion time
end For
Compute metric
Compute minimum completion time
Schedule the task
end While
for all tasks ti in meta-task Mv (in an arbitrary order)
for all hosts mj (in a fixed arbitrary order)
compute completion time
do until all tasks with high QoS request in Mv are mapped
for each task with high QoS in Mv, find a host in the QoS qualified
host set that obtains the earliest completion time
find the task tk with the minimum earliest completion time
assign task tk to the host ml that gives it the earliest completion time
delete task tk from Mv
update CTil for all i
end do
do until all tasks with low QoS request in Mv are mapped
for each task in Mv find the earliest completion time and the
corresponding host
find the task tk with the minimum earliest completion time
assign task tk to the host ml that gives it the earliest completion time
delete task tk from Mv
update CTil for all i
end do
The qos (priority based algorithm) finds the minimum earliest completion time and assign the task
to the host which gives the least completion time to it .Finally the low qos tasks are also mapped.
This algorithm improves efficiency by about 11%.

5. b Tree load balancing algorithms

The TLBA algorithm, named Tree Load Balancing Algorithm, creates a virtual interconnecting
tree (non-cyclic connected graph) among the computers of the system. On this tree, each computer
of an N level sends its updated load information to the N-1 level computers. The selection of the best
computer , to execute a process, received by the system, works as a deep search on the
interconnecting tree.

6. ARCHITECTURE:
We will study the architecture from the Open Grid Services Architecture (OGSA),
developed by the members of the Global Grid Forum (GGF).Building on existing Web
services standards, the OGSA defines a grid service as a Web service that conforms to
a particular set of conventions. Working groups in organizations like the Global Grid
Forum and OASIS are busy defining an array of grid standards in areas like
Applications and programming models, Architecture, Data management, Security,
Performance, Scheduling and resource management.
7 Grid Components:-
7.1P orta
l/Use r

Interface:-A grid user should not see all of the complexities of the computing grid. From this
perspective, the user sees the grid as a virtual computing resource just as the consumer of power
sees the receptacle as an interface to a virtual generator.
7.2 Security:- At the base of any grid environment, there must be mechanisms to provide
security, including authentication, authorization, data encryption, and so on. The Grid Security
Infrastructure (GSI) component of the Globus Toolkit provides robust security mechanisms.
The GSI includes an Open SSL implementation. It provides a single sign-on mechanism, so
that once a user is authenticated, a proxy certificate is created and used when performing actions
within the grid

7.3 Broker:-Once authenticated, the user will be launching an application. Based on the
application, and possibly on other parameters provided by the user, the next step is to identify the
available and appropriate resources to use within the grid. This task could be carried out by a
broker function
7.4 Scheduler:-Once the resources have been identified, the next logical step is to schedule the
individual jobs to run on them. If a set of stand-alone jobs are to be executed with no
interdependencies, then a specialized scheduler may not be required. However, if you want to
reserve a specific resource or ensure that different jobs within the application run concurrently,
then a job scheduler should be used to coordinate the execution of the jobs. The Globus Toolkit
does not include such a scheduler, but there are several schedulers available that have been
tested with and can be used in a Globus grid environment.
7.5 Data Management:-If any data including application modules must be moved or made
accessible to the nodes where an application's jobs will execute, then there needs to be a secure
and reliable method for moving files and data to various nodes within the grid. The Globus
Toolkit contains a data management component , Grid Access to Secondary Storage (GASS)
(facilities like Grid FTP)

7.6 Job and Resource Management:-. The Grid Resource Allocation Manager (GRAM)
provides the services to actually launch a job on a particular resource, check its status, and
retrieve its results when it is complete.

8. Grid Software Component:-

8.1 Management Component: There is a component that keeps track of the resources
available to the grid and which users are members of the grid.There are measurement
components and Advanced grid management softwares.
8.2 Donor Software:-It must be able to receive the executable file or select the proper one from
copies pre-installed on the donor machine. The software is executed and the output is sent back
to the requester.
8.3 Submission Software:- Usually any member machine of a grid can be used to submit jobs to
the grid and initiate grid queries. However, in some grid systems, this function is implemented as
a separate component installed on submission nodes or submission clients.

8.4 Schedulers:- This software locates a machine on which to run a grid job that has been
submitted by a user. Schedulers usually react to the immediate grid load and arranged
hierarchically.

8.5 Communications:- A grid system may include software to help jobs communicate with each
other However, the application may implement an algorithm that requires that the sub jobs
communicate some information among them.The open standard Message Passing Interface
(MPI) and any of several variations is often included as part of the grid system for just this kind
communication.

8.6 Observation, Management and Measurement:-This software is referred to as a sensor,

for implementing custom load sensors for other than CPU or storage resources. It forms the
basis for grid resource brokering, or to manage priorities more fairly.

9. Example Of Grid Computing:- United Devices Cancer Research

Project
The United Devices Cancer Research Project will research to uncover new cancer drugs
through the combination of chemistry, computers, specialized software, organizations and
individuals who are committed to fighting cancer. The research centers on proteins, which have
been determined to be a
possible target for cancer therapy, would go through a process called "virtual screening",
special analysis software (LigandFit) will identify molecules that interact with these proteins,
and will determine which of the molecular candidates has a high likelihood of being developed

into a drug.

The process is similar to finding the right key to open a special lock — by looking at millions
upon millions of molecular keys.
Participants in the United Devices Cancer Research Project (shown below) are sent a
ligand library over the Internet. Their PC will analyze the molecules using docking software
called LigandFit by Accelrys. The LigandFit software analyzes the molecular data by using a
three-dimensional model to attempt to interact with a protein binding site. When a ligand docks
successfully with a protein the resulting interaction is scored and the interactions that generate
the highest scores are recorded and filed for further evaluation.

10. Business & Technological Benefits From Grid Computing:

Increase productivity by providing users the resources they need on demand, respond
quickly to changing business and market demands, create virtual organizations that can
share resources and data, exploiting underutilized resources , parallel CPU capacity
,Resource balancing etc.

Characteristic Cluster Grid P2P

RESOURCEMANAGEMEN Centralized Distributed Distributed
T (I.E. MEMORY, OBJECTS,
STORAGE, NETWORK ACCESS,
ETC)
Singular Singular or multiple, Singular, multiple, or
RESOURCE OWNERSHIP (Often locked to a varies from platform distributed, depending
single node to to platform on circumstance and
prevent data architecture
corruption)
Method of Resource Centralized, Decentralized N/A, there is no single
Allocation / Scheduling allocated permanent host for
according centralized data or
configuration resource management.
Everything is transient.
Single Image Single or multiple Unknown, it is
EXTERNAL REPRESENTATION image(s) circumstantial

Guaranteed within Enforced within a Multiple competing

INTER-OPERABILITY a cluster framework standards
Mostly high-end, High-end or Any type, including
SUGGESTED EQUIPMENTS high capability commodity systems wireless device and
systems embedded systems.
2- 16 way Two to thousands Theoretically, infinite
SCALING (Although, units connection (In actuality, it depends
theoretically 128+ on network backbone
is possible) transmission speed,
number of clients, type
of transmission
protocol…..)
 Defined Centralized index, Always decentralized
DISCOVERY MECHANISM membership as well as, multiple discovery mechanism.
(Static or decentralized
Dynamic) mechanisms.

11. Limitations of Grid Computing:-

 Grid systems are less dynamic, scalable and fault tolerant as compared to P2P systems
 Not every application is suitable or enabled for running on a grid.
 Cellular wireless networks are more constrained than traditional wired networks because
of the limitations of bandwidth, processing power and memory.

12. Globus Grid Toolkit :-

The open source Globus Toolkit is a fundamental enabling technology for the "Grid”, which
includes software services and libraries for resource monitoring, discovery and management,
plus security and file management. In addition to being a central part of science and
engineering projects that total nearly a half-billion dollars internationally, the Globus Toolkit
is a substrate on which leading IT companies are building significant commercial Grid
products.

13. The Future:-Before grid computing moves into the commercial mainstream, CEOs need
to learn more about the technology and its possibilities, and identify ways they can use it. But
other problems needed to be solved like are security, standardization and new protocols for
bringing together a number of operating systems, vendor platforms and applications, before grid
computing becomes truly widespread—particularly in the context of inter-enterprise grids, utility
models and ultimately a global grid. A number of nonprofit groups such as the Global Grid
Forum, the Globus Project and the New Productivity Initiative are working on security and
standardization issues.
14. Conclusion:-So we see, grid computing *is* the ultimate "killer” technology. The grid
computing can be implemented using different technologies like XML, java, CORBA and
dynamically allocates the resources. The basic part is the scheduler of the grid computing. The
grid can be implemented in intranet or on Internet. The grid is used for large computations in fields
such as biotechnology, automobile, medical and many fields. The grid can be implemented
depending on requirements of the application, security, and priority for task to be done. The
scheduler is the core item of the grid; the algorithm can be depending on priority or resources
available or type of network available. Although there is some limitation on wireless environment,
but it is future of computing which will explore new ways for computing in enterprise environment.