Grid computing

Grid Computing

By:Neha Bhambu

00CE035B.Tech Final yr

Computer Science

Seminar report

Introduction

SETI@Home made the idea of finding extraterrestrial life exciting by allowing the public to process radio logs for peaks and regular signals by providing the computing power of a supercomputer by utilizing scavenging CPU cycles of so many idle PCs disperserd

at so many diverse locations all around the world . Distributed.net concentrates on the computational needs of solving encrypted strings through brute force methods .

NSPA ,Globus grid project ,Teragrid ,zetagrid are providing viable grid services through the use of Web

Services based on OGSA/OGSI standards. A virtual organization is a collaboration of various

individuals or institutions with a common goal sharing resources like storage , computing power , software, data ,file sharing in a highly controlled way with clearly defined access rights .

A Grid is a collection of distributed heterogeneous computing and data resources available through network tools and protocols which enable controlled and coordinated sharing and appear as a single large virtual computing system

Types of grid :

1.On the basis of shared resources:

Computational grid

Data Grid

Scavenging grid

2.On the basis of

topology :

Cluster grid

Intra grid

Inter grid

Extra grid

3. On the basis of

behavior:

Enterprise grid

Partner grid

Service grid

Grid Layered Structure

OGSA

• OGSA is a distributed interaction and computing architecture that is based around the Grid service, assuring interoperability on heterogeneous systems so that different types of systems can communicate and share information.It is itself evolving and attempts to describe the core capabilities of the grid. definition of standard service interfaces and protocols.

• Some of the facilitites provided:• Naming: Each grid service instance is globally, uniquely,

and for all time named by a Grid Service Handle (GSH).• Factories: Create new grid service instances and

maintain a group of service data elements that can be queried. A factory will also have an associated registry to keep track of these instances and enable discovery. s.

• Stateful: A grid service instance has a state. A process can be initiated via a method call on a service port type and its state checked using the GSR.

OGSI• The Open Grid Services Interface (OGSI) specification

is a• companion standard that defines the interfaces and

protocols that will be used between the various services in a grid environment. The OGSI is the standard that will provide the inter-operability between grids designed using OGSA.

• OGSI consists of specifications on how work is managed, distributed, and how service providers and grid services are described. Web services, particularly the Simple Object Access Protocol (SOAP) and Web Services Description Language (WSDL), are a major part of this specification.

Grid Service Semantics• Web Services model used as it is a distributed object

model. • The Grid service interface is described by WSDL to

use the service. A new tag gsdl has been added to the WSDL document for Grid service description. The UDDI registry and WSIL document are used to locate Grid service. The transport protocol SOAP is used to connect data and application for accessing Grid service .

• The interfaces of Grid services address discovery, dynamic service instance creation, life-time management, notification, and manageability.

Grid Service Handling• Service versioning and upgrading

The portTypes of a Grid service and the versioning information are specified in the Grid service's serviceType in OGSA. Grid Services in a distributed system can be independently upgraded based on the versioning information defined in WSDL. The compatibility between Grid services can be managed in OGSA based on the versioning information. The service client can easily find and invoke a Grid service with right method signatures.

• Grid service instance creation and invocation

The user application invokes create Grid service requests on the factory interface to create a new service instance. An initial lifetime of the instance can be specified before the service instance is created.The newly created service instance will keep the user credentials for interacting with other systems over the Internet. The newly created Grid service instance will be automatically assigned a globally unique name called the Grid Service Handle (GSH). In a transient stateful service if a failed operation happens, the keeplive messages stop . Then the Grid service instance automatically times out .

Grid Service Commands

Grid Service Deployment

• The Remote Procedure Call (RPC) servlet of Simple Object Access Protocol (SOAP) and the real implementation of the Grid service can be deployed on

an application server . All the invocation messages will be captured by the SOAP RPC servlet, which routes the messages to the corresponding Grid service. Grid services can be published to a Universal Description, Discovery and Integration (UDDI )egistry, or WS-Inspection (WSIL ) documents .

GRAM GRAM is the module that provides the remote execution and

status management of the execution. When a job is submitted by a client, the request is sent to the remote host and handled by the gatekeeper daemon located in the remote host. Then the gatekeeper creates a job manager to start and monitor the job. When the job is finished, the job manager sends the status information back to the client and terminates.

MDS(Monitoring and Discovery Service)

MDS gives users the ability to obtain vital information about the grid and grid resources. It utilizes LDAP(lightweight directory access protocol) to execute queries. MDS provides access to static and dynamic information of resources. Basically, it contains the following components: Grid Resource Information Service (GRIS) Grid Index Information Service (GIIS) Information Provider MDS client

GSI• The infrastructure is based on the SSL protocol (Secure

Socket Layer), public key encryption, and x.509 certificates.

• GSI authenticationGSI offers users a secure authentication option by creating a time-stamped proxy based on the user's private key. Users can't submit jobs to run or transfer data without creating the proxy. Once created, the proxy is used to grant -- or deny -- access to resources found throughout the grid. Because the proxy is used across the system, this gives the end user the ability to sign on only once. In a PKI, all entities (users and resources) are identified by a globally unique name known as a Distinguished Name (DN).

• GSI authorizationThe GSI handles user authorization by mapping the user to a local user on the system being accessed. In a GSI-enabled grid, the system receiving the request reads the user's name from the proxy, and then accesses a local file to map that name to a local user.

• GSI secure communicationBy default, GSI secures communications using digital certificates for mutual authentication and SSL/TLS for data encryption. The Toolkit contains OpenSSL, which it uses create an encrypted tunnel between grid clients and servers.

• A proxy credential is a short-term credential that is created by a user,which can be used in place of the long-term credential to authenticatethat user .

Grid FTPGridFTP provides a secure and reliable data transfer among grid nodes. The word GridFTP can referred to a protocol, a server, or a set of tools. GridFTP protocolGridFTP is a protocol intended to be used in all data transfers on the grid.GridFTP server and clientGlobus Toolkit provides the GridFTP server and GridFTP client, which are implemented by the in.ftpd daemon and by the globus-url-copy command, respectively. They support most of the features defined on the GridFTP protocol.The GridFTP server and client support two types of file transfer: standard and third-party. The standard file transfer is where a client sends the local file to the remote machine, which runs the FTP server. Third-party file transfer is where there is a large file in remote storage and theclient wants to copy it to another remote server.

Grid service working :

Grid Computing Environment

• This describes the “user side” of a computing system where users interact via which controls a set of distributed back end resources. GCE’s fulfill (at least) two functions :

• “Programming the User Side of the Grid”• Controlling user interaction – rendering any output and

allowing user input in some (web) page. This includes aggregation of multiple data sources in a single portal page.

• A nugget is defined as a code module individually programmed to provide a particular service like SQL interface . GCE programming refers to their integration.This integration is termed as workflow and in acommon modelled paradigm a graphical interface where one can choose nuggets from a palette and link ‘ports’ or channels of the nugget.

GCE Shell

• Catalog of primitive functions needed to program the grid enabled as a command line interface .

GCE Shell must express:• Negotiated interaction

between nuggets and users.

• Files and services hirarchies at all levels of system .

• Distinction between object and its metadata.

GCE Shell must have following features:

• Search• Discovery• Registration• Security• Better workflow• Groups and

collaboration features • Meta-data handling• Management and

Scheduling• Networks• Negotiation primitives

for service interaction

Problem Solving Environments

• A problem solving environment is a computational system that provides a complete and convenient set of high level tools for solving problems from a specific domain. The PSE allows users to define and modify problems, choose solution strategies, interact with and manage appropriate hardware and software resources, visualize and analyze results, and record and coordinate extended problem solving tasks.

User Portal with multiple grid services

SandBox

Grid Bandwidth Management

Bandwidth considerations comprise :

CPU usage The CPU bandwidth of the grid is the maximum rate at which the grid can operate.

Network usage. The bandwidth available to

a grid and that used by the grid in its normal operation. Larger bandwidth availability enables us to build more complex, high power grids.

• Bandwidth will be affected by a number of factors :

• Number of clients in the grid • Number of work units distributed

each time • Period required to process each

unit • Work unit size (source and

response)• A WorkQueue table holds

the information about all of the work units in the system and their status, such as waiting, assigned to a client, or completed. A second table, ActiveUnits, holds the information about which clients have which units.

Cooperative computing using a dynamic server.Each client can measure its own performance and then report it to broker.

Job Flow• Serial job flowIn this case there is a single

thread of job execution where each of the subsequent jobs has to wait for its predecessor to end and deliver output data as input to the next job. This means any job is a consumer of its predecessor, the data producer.

Parallel job FlowEach job may receive a

discrete set

of data, and fulfills its computational task independently and delivers

its output.Network job flowMore analysis needs to be

done to determine how best to split the application into individual jobs, maximizing parallelism. It also adds more dependencies on the grid infrastructure services such as schedulers and brokers .

Conclusion• Grids have become viable by the use of

standards like OGSA and OGSI and are instrumental in harnessing huge computing power and storage resources by coordinated and controlled sharing on heterogeneous systems.

• In future grid computing will become ubiquitous and use of web services will become even more pronounced as web services are based on distributed computing model .

• Grid services are transient ,autonomus and are being developed as large inter grids .

• Grid services are being sponsored by industrial and techno bigwigs like IBM, Entropia, NSPA and standards being constructed at Global Grid Forum .

• Globus toolkit versions are being developed which provide features for grid services development.

• Security is a big consideration and GSI norms enforcement assures reliability.

ReferencesIBM Redbooks:

Louis Ferriera , Victor Bersotis. 2002. Introduction to Grid Computing with Globus

Bart Jacobs, Norbert Breinstein 2002. Enabling Applications for Grid Computing with globus

White Papers

Ian Foster Carl Kesselman Steven Tuecke 2001 The Anatomy of the Grid

Ian Foster Carl Kesselman Steven Tuecke 2002

The Physiology of the Grid

AIan Foster Carl Kesselman Steven Tuecke 2002

Security Architecture for Grids

Geoffery Fox 2001

Grid Computing Environments

Gregor von Laszewski et.al 2000. Problem Solving Environments and Portals

Web Pages:

Qun Zhou Leng Zheing 2002 . Developing Grid computing applications, Part 1_files, Part2_files.html

Martin C.Brown 2003 .Merging grids and web services.html