Transcript Virtuoso_plab - Computer Science Division

Virtuoso: Resource Management and Prediction
for Distributed Computing Using Virtual Machines
Peter Dinda
Ashish Gupta
Ananth I. Sundararaj
Department of Computer Science
Northwestern University
Outline
Objective
 Background
 Proposed Idea
 Research Issues
 Implementation Status
 Demo
 Conclusions

Objective

Propose a new paradigm for Grid computing



Operating System level Virtual Machines (VM)
Middleware mechanisms to manage VMs in a
distributed environments
To demonstrate the feasibility of the
proposed idea

Quantitative results
• Functionality perspective
• Performance perspective
Background

Grid Computing
Fundamental Goal: Seamlessly multiplex
the distributed computational resources
of providers among users across a wide
area network
 Current Approach: Resources are
multiplexed using the mechanisms found
in typical operating systems

Proposal: Raise the level of Abstraction
OS Virtual Machine
OS User
Virtual Machines

History


First came about in the 1960's on mainframes
as a way to create less complex multi user
time share environments
What is it?

A virtual machine is an abstraction of a
physical machine
• Created using a Virtual Machine Monitor (VMM)
running on a physical machine

Gives the illusion of working on a separate
machine
Virtual machines contd..

Architecture

The abstraction of a virtual machine is that each user
appears to have a dedicated machine at their disposal,
the hardware of which they can access directly
VMware


VMware, Inc. was
founded in 1998
Currently the
largest producer
of virtual machine
software
Proposed Idea


To use virtual machines in grid computing
Problems with traditional approach to grid
computing ( OS User approach)




Security
Site dependence
Migration of user tasks
Virtual machines address the above
mentioned problems
Proposed Idea contd..

Other Advantages of using Virtual
Machines
Isolation
 Customization
 Legacy Support
 Administrator Privileges
 Resource Control

Research Issues
Architecture
 Virtual Networking
 Performance
 Data Management
 Resource Management
 Integration with existing Grid
Infrastructure

Architecture

Typical Components include
User
 Information server
 Physical server
 Image server
 Data server


These components are distributed
across the nodes of the grid
Physical Servers
Image Server
Data Server
User Front End
Information Service
Framework

To build a virtual computational grid
support framework



Virtual machine instances and capability for
instantiating virtual machines are advertised
via a grid information service
A user consults it to locate a physical machine
to instantiate a virtual machine on
The user may also consults the information
service to locate virtual image server with a
base OS installation or may decide to
customize one
Framework Cont…




The user can now instantiate a virtual
machine on the selected physical machine
An operating system can then be installed
using a virtual CD-ROM and the virtual
machine can be customized
The user (owner of the virtual machine) could
negotiate terms with the physical server
Virtual networking would put the virtual
machine on the user’s network
Current Work

Create an infrastructure for
experimenting with our ideas

Initial Part:

A menu driven application which
provides a framework for grid
computing using virtual machines
Perl Scripts
Current Goals
Remote Instantiation
 Remote Display
 Virtual Network of Virtual Machines
 Remote CD-ROM access
 Migration of Virtual Machines

User should be able to say:
‘ssh root@vm-8-1`
Remote Display

Requirements


Remote Display options




A dedicated console for the remote virtual
machine
X-Server
VNC
Windows Remote Desktop
Issues


Interactivity
Security
Virtual Networking

Requirement
A virtual network which connects the VMs
instantiated by the user
 Should appear as a part of the user local
network

Integration into Network

Case I



Physical server (VM host) is VM friendly
VM host provides IP addresses to dynamic VM instances
Case II




VM host does not provide IP addresses to dynamic VM
instances
Traffic is tunneled through at the Ethernet level,
between the remote virtual machine and the local
network of the user
Remote machine would appear to be connected to the
local network
We assume that we are able to establish a TCP
connection to the remote site, in particular the remote
site only allows SSH traffic
Issues
Ethernet Level Tunneling and Routing
 Performance of virtual network
 Overlay network Optimization

User and application (parallel
applications) requirements
 Efficient Communication


Adapt to Migration of virtual machines
VMWare Networking Options

Bridged Mode (Ist case)


Just like any other machine on the network
Host Only Networking (2nd case)

Host-only networking creates a network that is completely contained
within the host computer
Current Implementation

Framework



Each physical server will run a daemon called
the bridge daemon
Bridge daemon tunnels traffic via a ssh tunnel
from remote network to virtual network
adapter
The daemons can communicate with each
other using a text based protocol and
reconfigure themselves to establish desired
network configurations
Resource Management


The use of Virtual machines create new
challenges for scheduling and resource
management
Resource perspective




Virtual machines provide a mechanism for controlling
how and when resources are used
Resource owners will be more conducive to allowing
their resources to be used
The user sees a raw machine completely at their
disposal, while the resource owner only sees a single
entity to be scheduled
A real time schedule will be required to accommodate
the resource owner’s constraints and user needs
Resource Management contd..

Application perspective
Applications have to adapt to the
dynamically available resources
 Virtual machines simplify this as they
can carry the preferred execution
environment
 Virtual machines are themselves a
resource and need to be registered and
advertised

Data Management

This involves

Transfer of VM images such that they
can be instantiated anywhere and
migrated when necessary
• On demand transfer
• Seamless migration while keeping remote
data connections active

Support for location independent
access to user files
• Move the computation closer to the data
Conclusions
Provided an argument for the use of
virtual machines in grid computing
 Detailed research issues involved
 Explained the current implemented
framework
