By Mohd. Ajaz Ahmed

GRID VIRTUALIZATION

ENGINE

Introduction

Virtual Machine Based Grid System

Architecture Of GVE

Use Cases Of GVE

Queries

AGENDA

Grid provides for the allocation and management of heterogeneous networked resources and makes them available to a large number of users.

A network of geographically distributed resourcesResources may be owned by diverse organizations.Grid Computing involves more than one computer to

solve the problem.Grid Computing is used mainly for CPU intensive

calculations which can be divided into multiple tiny tasks and which can be executed parallel on different nodes.

LET’S LEARN ABOUT GRID

LET’S LEARN ABOUT GRID

Control Server

TASK

Grid node

Virtualization, in computing, refers the act of creating a virtual version of something, including but not limited to a virtual  computer hardware platform, operating system(OS), storage device, or computer network resources.

In computing, virtualization means to create a virtual version of a device or resource, such as a server, storage device, network or even an operating system where the framework divides the resource into one or more execution environments.

Devices, applications and human users are able to interact with the virtual resource as if it were a real single logical resource. 

LET’S LEARN ABOUT VIRTUALIZATION

LET’S LEARN ABOUT VIRTUALIZATION

Although great advances have been made in the field of Grid computing, users still expect to meet some difficulties when employing Grid resources. Qualities of service (QoS) of resource provision and performance Isolation and Customized runtime environment for Grid applications

Goal of a Grid infrastructure is to provide “flexible, secure, coordinated resource sharing.

The middleware of these approaches interact with physical resources at the same level as local users and applications do.

Virtualization technologies encompass a variety of mechanisms and techniques used to decouple the architecture and user-perceived behavior of hardware and software resources from their physical implementation.

OBJECTIVE

Lower Hardware Cost

Easier Management

Isolation between

Logical servers

BENEFITS OF VM APPROACH

Grid Virtualization Engine creates an abstract layer between users and underlying virtualization technologies.

Can also be simplified as software layer between various virtualization implementations, computing centers and Grid users.

Enables Users to operate VM’s and provide Virtual Distributed environments

Users can require and employ virtual machines via the access interface of GVE

GRID VIRTUALIZATION ENGINE

Access ServiceVirtual MachineHost Resource

Computational Grid Computer Site

allowsremote users to access resources of the computerCenter

Form virtualized distributed environments. GVE service operates on virtual machines in this level Host resources offer

multiple virtual machines. The GVE Agent Service is implemented

GRID VIRTUALIZATION ENGINE

ARCHITECTURE OF GVE

GVE provides functionalities for users to access virtual machines and virtual environments supported by distributed computing centers.

The GVE is designed in distributed and hierarchical flavors with standard Web service interface.

Current implementation of the GVE can work on popular VMMs, for instance, VMware server, and VMware ESX server.

GVE contains the following components: GVE Site Service, GVE Agent Service, and Virtual Machine Disk Database.

GVE Web Service

User Info. Service

User Info. DatabaseGVE Site

Service

VM Image Database

Registry Database

Agent Web Service

Registry ServiceVMM

GVE AGENT

ARCHITECTURE OF GVE

Works on a computer center’s access point.Manages host resources inside the center by

communicating with the GVE Agent Services that run on the host resources.

The GVE Site Service includes three main components: the GVE Web Service, the User Information Service and the User Information Database

GVE SITE SERVICE

The GVE Web Service itself is the component which is responsible of the business logic of the GVE Site Service.

It decides to which virtual machine requests should be sent to and defines the policies of resource allocation.

Web service technologies provide a framework for developing distributed applications that run on the Internet

For further purpose it needs the User Information Service

GVE WEB SERVICE

GVE Site Service has to store data that may be concurrently used by other components in the GVE

Grid technologies enable large-scale sharing of resources within formal or informal consortia of individuals and/or institutions: what are sometimes called virtual organizations.

In these settings, the discovery, characterization, and monitoring of resources, services, and computations can be challenging due to the considerable diversity, large numbers, dynamic behavior, and geographical distribution of the entities in which a user might be interested.

Hence, information services are a vital part of any Grid software infrastructure, providing fundamental mechanisms for discovery and monitoring, and thus for planning and adapting application behavior.

USER INFORMATION SERVICE

The User Information Database records the management policies and accounting information of virtual machine usage

To check whether the resource quota of the user has been reached before new resources are allocated

It can control how long a resource has been granted to a user

USER INFORMATION DATABASE

Runs on the host resource Receives operation commands from the GVE Site Service

and talks with the specific VMM Virtualization technology dependent For each type of VMM, a corresponding GVE Agent Service

should be implemented GVE Agent Service knows how to call management

functions of the underlying VMMs VMM is managed by a GVE Agent Service It provides VM from underlying VMMs to a GVE Site Service Registry Service provides the following functions to access

the Registry Database, which stores the state of virtual machines, the GVE Site information, the state of the virtual machine request and management.

GVE AGENT SERVICE

Might not be directly implemented by the GVE developer

Stores the data which are needed for VM creation and management

VM DISK DATABASE

USE CASES OF THE GVE

Request new VM

Request

Owned VM

Request VM

Manage VM

Compute

allocated

resourceCheck Identit

y

Release VM

Manage GVE

GCSUser Info. Service

GVE Administrator

Actors of this use case are the GCS and the User Information Database

Provision of VM based on demand and requirements of the user

GVE user provides the profile of the virtual machine he requires Existing free virtual machine or creates a new VM to fulfill the user requirement

User is registered to the GVE

REQUEST A VM

Actors are the GCS and the User Information Database.

GVE provides users with functions to manage the virtual machines that they have acquired, for instances, start, stop, suspend, resume, and migrate virtual machines.

Requires either the GCS has acquired a virtual machine or the GCS has provided details about the virtual machine to be managed.

Triggered when the GCS sends a request of managing virtual machines.

GCS receives a message notifying the success of the operation or failure of the operation

MANAGE A VIRTUAL MACHINE

Actors of this use case are the GCS and the User Information Database.

When a GVE user has finished a computational task on the virtual machine and does not need it anymore, the user releases a virtual machine.

In case the allocated time for virtual machine usage has passed, the GVE will stop the usage of the virtual machine.

Before this use case is triggered, the GCS should provide details about an acquired virtual machine.

After this use case finishes, the virtual machine is stopped and its profile is changed.

RELEASE A VIRTUAL MACHINES

Actor of this use case is the GCS. Resources allocable to a GVE user should not exceed the

predefined resource quota for the user. Used to compute resources for user’s allocation. Each time a user requests a new virtual machine, the total

amount of resources that are allocated to the user at that time is computed to check whether it exceeds the quota for the user.

It demands information of the user information and the virtual machine profile.

If this use case successfully ends, the GVE has permissions to allocate one resource with the given profile to the user.

Otherwise, the resource quota for the user has been reached then the GVE does not have the permission to allocate resources with the given profile to the user.

COMPUTE ALLOCABLE RESOURCES

Actor of this use case is the User Information Database.

The GVE identifies all users before they can perform any operation on the GVE.

The identity is also needed to know whom a virtual machine belongs to and how many resources a user can acquire.

It requires that the user has appropriate proof of his identity.

It’s successful execution means the user is verified. Otherwise, the user is not approved.

CHECK IDENTITIES

The correct setup of the GVE requires some pre-configurations by an administrator.

As a GVE is a distributed application, the administrator identifies all components that build a new GVE and connect them together.

The Administrator also manages the identity of the different entities that use the GVE. It includes registering or unregistering of a Virtualization Agent and registering or unregistering a user.

MANAGE THE GVE

THANK YOU