M.Tech. [wcc] 2003
The last decade has seen a substantial increase in commodity
computer and network performance, mainly as a result of faster
hardware and more sophisticated software. Nevertheless, there
are still problems, in the fields of science, engineering, and
business, which cannot be effectively dealt with using the current
generation of supercomputers. In fact, due to their size and
complexity, these problems are often very numerically and data
intensive and consequently require a variety of heterogeneous
resources that are not available on a single machine or in single
organization. These two factors combines and leading to the
possibility of using distributed computers as a single, unified
computing resource, what is popularly known as Grid computing.
Grid is a type of parallel and distributed system that enables
the sharing, selection, and aggregation of geographically distributed
autonomous resources (owned by different organizations) dynamically
at runtime depending upon their availability, capability, performance,
cost and users quality of service requirement.
A high-level view of activities within the Grid is shown in
The users interact with the Grid resource broker to solve problems,
which in turn performs resource discovery, scheduling, and the
processing of application jobs on the distributed Grid resources.
To build a Grid, the development and deployment of a number
of services is required. These include security, information,
directory, resource allocation, and payment mechanisms in an
open environment and high-level services for application development,
execution management, resource aggregation, and scheduling.
2. Virtual Organizations
Software tools and services providing the capabilities of grid
to link computing capability and data sources in order to support
distributed analysis and collaboration are collectively known
as Grid middleware. As Grid computing provide user with a seamless
computing environment, the Grid middleware system needs to handle
several challenges. Some of them are:
Multiple administrative domains and autonomy : Grid
resources are geographically distributed across multiple administrative
domains and owned by different organizations. The autonomy of
resource owners needs to be honored along with their local resource
management and usage policies.
Dynamic Nature : In a Grid, resource failure is the
rule rather than the exception. In fact, with so many resources
in a Grid, the probability of some resource failing is high.
Resource managers or applications must tailor their behavior
dynamically and use the available resources and services efficiently
Heterogeneity : A Grid involves a multiplicity of
resources that are heterogeneous in nature and will encompass
a vast range of technologies.
Scalability : A Grid might grow from a few integrated
resources to millions. This raises the problem of potential
performance degradation as the size of Grids increases.
To tackle these challenges, Grid architecture has been proposed
of the creation of Virtual organizations (VO's) by different
physical organization coming together to share resource and
collaborating in order to achieve a common goal. A VO defines
the resources available for participants and the rules for accessing
those resources. Within a VO, participants belonging to member
organizations are allocated share based on urgency and priority
of a request as determined by the objective of VO.
3. Grid Components
In an World-wide Grid environment,
capabilities that the infrastructure needs to supports includes:
- Remote Storage and replication of data sets
- Publication of data sets using global logical name
- Security- Access authorization and uniform authentication
- Uniform access to remote resource (Data and computational
- Publication of services and access cost
- Discovery mechanism for suitable Datasets and computational
- Mapping and Scheduling of jobs (Aggregation of distributed
- Submission and Monitoring of job execution
- Movement of data between the user machines and distributed
- Enforcement of QOS requirements
- Metering and Accounting of resource usage
These capabilities in Grid computing environment play a significant
role in variety of scientific, engineering and business applications.
Various grid components providing these different capabilities
are arranged into layers. Each layer builds on the services
provided by lower layer in addition to interacting and co-operating
with components at the same level. It consist of four layers
: fabric, core middleware, user middleware, application and
Grid fabric : This consists of all the globally distributed
resources that are accessible from
anywhere on the Internet. These resources could be computers
(such as PCs or Symmetric Multi-Processors) running a variety
of operating systems (such as UNIX or Windows), storage devices,
Databases and special scientific instruments such as a radio
telescope or particular heat sensor.
A layered Grid architecture and components
Core Grid middleware : This offers core services such
as remote process management, co-allocation of resources, storage
access, information registration and discovery, security, and
aspects of Quality of Service (QoS) such as resource reservation
and trading. These services abstract the complexity and heterogeneity
of the fabric level by providing consistent method for access
User-level Grid middleware : This includes application
development environments, programming tools, and resource brokers
for managing resources and scheduling application tasks for
execution on global resources. It utilizes the interfaces provided
by low level middleware to provide higher level abstractions
Grid applications and portals : Grid applications
are typically developed using Grid-enabled languages and utilities.
An example application, such as parameter simulation or a grand-challenge
problem, would require computational power, access to remote
data sets, and may need to interact with scientific instruments.
Grid portals offer Web-enabled application services, where users
can submit and collect results for their jobs on remote resources
through the Web.
4. Operational flow
Enabling the resource constituents of the Grid, they need to
be accessible from different management domains. This can be
achieved by installing Globus in Unix/Linux environment or Alchemi
in Windows environment as a core Grid middleware services. Multi-node
resources can be abstracted to single resource to the Grid by
using job management system such as Condor, PBS etc. Data grid
technologies like SRB, Globus RLS or EU Data grid may be deployed
for the environments where databases to be federated for sharing
among various parties.
The following steps show the key steps involved in interaction
between various grid components from user's perspective.
- Users compose their application as distributed application
using development tools.
- Specifying QOS parameters and submit job to Grid resource
- Broker performs resource discovery and their characteristics
using Grid information service.
- Broker identifies resource prices by querying Grid market
- Broker list Datasets or replicas and selecting optimal ones.
- It also lists the computation resources that provide required
- Broker interacts with user to check out necessary credit
- The broker scheduler maps and deploys data analysis jobs
on resources that meet user QOS requirements.
- The broker agent on a resource executes the jobs and returns
- The grid resource broker collates the results and passes
- The metering system charges the user by giving resource
usage information to the accounting system.
Grid Computing is becoming the preferred platform
for next generation eScience experiments that require management
of massive distributed data. Presented article covers the basic
overview architecture for understanding of Grid computing. It
includes software-layered architecture where different components
providing different services configured to different layers
depending upon their functions. All these combines to provide
a seamless computing environment where user just need to submit
jobs to Grid.