Performance Test of Virtual Linux Desktop Cloud Services on System z
An IBM Redpaper publication
Published 05 February 2010
IBM Form #: REDP-4593-00
Authors: Mike Ebbers, Tanya S Buchanan, Dinakaran Joseph, Andrea Greggo, John Langer, Eugene Ong, Mark Wisniewski
Cloud computing is an evolutionary deployment model for IT and business assets. For the user, cloud computing delivers flexible business models by providing consumable and affordable services that are invoked and accessed through self-service Web 2.0 portals. For the provider, cloud computing enables the delivery of IT and business assets that are operationally efficient and highly virtualized. Across industries, cloud computing is providing a fast, simple, and inexpensive means to provision computing resources for a variety of key applications, platforms, and infrastructure services.
Focusing on poorly performing business and technology processes enables companies to begin transforming to cloud computing while targeting near-term return on investment. Managing and provisioning end user desktops is an example of a costly and inefficient business process that can be greatly improved with cloud computing alternatives. Centrally controlling and managing end user desktop environments and making them available as a cloud service is a great way to reduce support costs and increase end user productivity.
This IBM® Redpaper™ document describes the results of testing a virtual Linux® desktop service that was created using eyeos, an open source Web desktop, running on an IBM System z® server. This configuration was then tested by creating thousands of virtual clients concurrently accessing the desktop platform on Linux on System z to determine how a shared server environment would respond to concurrent desktop demands.
The results of the desktop platform configuration test validate that a shared all-Linux platform running z/VM® virtualization is an excellent environment for efficiently using CPU, memory, and I/O to manage thousands of concurrent desktop users.
Table of contents
Others who read this publication also read
Follow IBM Redbooks
Follow IBM Redbooks