Introduction to Grid Computing

1
Introduction to Grid Computing

• Presented by Marty ONeill II

2
Cluster Computing as compared to Grid Computing

• Clusters are constructed to solve the problem of
  inadequate computing power.
• They are limited in scope and have dedicated
  functionality.
• Cluster nodes are centrally controlled and the
  cluster manager is aware of the state of each
  node.
• The concepts used in cluster computing are only a
  subset of that which is used in grid computing.

3
The Worlds Largest Computer
Grid Computing involves connecting a potentially
unlimited number of computers into a grid.
This grid will provide computational power
on-demand to users much like the power grid
provides electricity to users at a flick of the
switch.
Grid Computing is an effort to build the
computational infrastructure needed to satisfy
the increasingly complex scientific and
industrial problems of both today and tomorrow.
4
Requirements for Grid Computing
Decentralized control infrastructure to provide
coordination among resources is needed
The ability to manage all aspects of data (e.g.
data location, transfer, access)
Open standard protocols and frameworks this
allows for interoperability and integration
facilities
Failure detection and failover mechanisms
Ability to provide mechanisms that can
intelligently and transparently select computing
resources capable of running a users job
Ability to meet QoS requirements response time
measures, aggregated performance, security,
resource scalability, availability
5
Virtual Organizations and Resource Sharing
Software Application Provider
Weather Prediction
Problem A
Dynamic Virtual Organization A formed to provide
weather prediction
Weather prediction
Math Modeling Application
Hardware Service Provider
User
Computer Cluster

• Virtual organizations are
• Logical entities
• Dynamically created to solve a specific problem
• On-demand resource allocation and provisioning
  for solving the problem

Network Bandwith
Blades

resources
User
Financial Expert Service Provider
Dynamic Virtual Organization B formed to solve a
financial modeling
Financial Modeling
Problem B
Financial Modeling
Database System
6
Requirements for Users and Virtual Organizations

• Users must be able to
• Clearly and unambiguously identify the problem to
  be solved
• Identify and map the resources required to solve
  the problem
• Meet required QoS while not violating SLAs
• Collect feedback regarding resource status

• Virtual Organizations must be able to
• Form virtual task forces to solve specific
  problems
• Dynamically collect resources from different
  providers based upon the users needs and the
  problem
• Identify problems, linking them to required
  resources or service providers, and automatically
  resolve them
• Dynamically provision and manage the resources
  while not violating SLAs

7
Grid Computing Organizations
The most prominent of these is the Global Grid
Forum (GGF). It is a means of coordinating Grid
computing technology efforts, promoting reuse and
interoperability, and sharing the results. It
includes more than 400 organizations such as
research institutions, universities, and
commercial organizations from all over the world.

Grid Users (NSF, TerraGrid)
Corporate Toolkits and Users (IBM, Avaki/Sun, HP,
MS, )

Toolkits and middleware solution
providers (Globus, OGSI, NET, NMI)
Standards Organizations (GGF, IETF, DMTF, W3C,
OASIS)
8
Goals of the Global Grid Forum
Create an open process for the development of
grid agreements and specifications.
Handle intellectual property policies.
Provide a forum for information exchange and
collaboration.
Create grid specifications, architecture
documents, and best practice guidelines from the
experience of the technologies associated with
grid computing.
Improve collaboration among the people involved
with grid research, grid framework builders, grid
deployment, and grid users.
Manage and version control the documents and
specifications.
Educate on advances in the grid technologies and
share experiences among the people of interest.
9
Grid Computing Organizations
The most prominent of these is the Global Grid
Forum (GGF). It is a means of coordinating Grid
computing technology efforts, promoting reuse and
interoperability, and sharing the results. It
includes more than 400 organizations such as
research institutions, universities, and
commercial organizations from all over the world.
Globus is one of the most prominent organizations
responsible for the toolkits, middleware, and
framework for grid computing. It is a
multi-institutional research effort to create a
basic infrastructure and high-level services for
a computational grid.

Grid Users (NSF, TerraGrid)
Corporate Toolkits and Users (IBM, Avaki/Sun, HP,
MS, )

Toolkits and middleware solution
providers (Globus, OGSI, NET, NMI)
Standards Organizations (GGF, IETF, DMTF, W3C,
OASIS)
10
Globus Toolkit
The Globus Toolkit was conceived to remove
obstacles that prevent seamless collaboration.
GT has grown through an open-source strategy,
distinct from proprietary attempts at
resource-sharing software.
Globus Toolkit 3 (GT3) is currently the latest
version released. GT4 is still under production.
GT3 includes software for security, information
infrastructure, resource management, data
management, communication, fault detection, and
portability.
GT3 is the first full-scale implementation of new
Open Grid Services Architecture (OGSA)
specifications defined (in a large part) by
Globus.
11
Problems in Grid Computing
Grid Computing
Coordination

• Sharing Resources
• Computing Power
• Data Storage
• Hardware Instruments
• On-Demand Software Applications

High-Performance Orientation
Event Correlation
Authentication
Authorization
Access Mechanisms
Resource Discovery
This all becomes more complicated when the grid
is designed to be a utility!
12
More on Virtual Organizations
They are the key to grid computing.
Assigning users, resources, and organizations
from different domains across multiple, worldwide
geographic territories to a virtual organization
is one of the fundamental technical challenges in
grid computing.
Virtual Organizations are sets of users defined
around a set of resource-sharing rules and
conditions.
Remember
Different Virtual Organizations may vary from
each other in terms of size, scope, duration,
sociology, and/or structure.
13
Grid Architecture
The Resource Layer controls the secure
negotiation, initiation, monitoring, metering,
accounting, and payment involving the sharing of
operations across individual resources.
The Connectivity Layer defines the core
communication and authentication protocols
required for grid-specific networking services
transactions.
The Collective Layer is responsible for all
global resource management and interaction with a
collection of resources.
The Application Layer is where the user-defined
grid applications reside.
The Fabric layer defines the resources that can
be shared.
Application
Application
Collective
Internet Protocol Architecture
GRID Protocol Architecture
Resource
Transport
Connectivity
Internet
Fabric
Link
14
References

• Berman, F., Fox, G.C. Hey, A.J.G. Grid
  Computing Making the Global Infrastructure a
  Reality. San Francisco, CA Wiley, 2003.
• Foster, I. Kesselman, C. Grid 2. San
  Francisco, CA Morgan Kaufmann Publishers, 2004.
• Getov, V. et. Al. Performance Analysis and Grid
  Computing. USA Kluwer Academic Publishers Group,
  2004.
• http//www.globus.org
• Joseph, J. Fellenstein, C. Grid Computing.
  Upper Saddle River, NJ Pearson Education, 2004.