Skip to content

High I/O + Real-time + Virtualization == Guest OS’s may not get enough resources

April 27, 2010

xenCloud computing often uses server virtualization on its back-end, implementing many operating systems on a single piece of hardware. In order to share these resources, software called a hypervisor must exist, and methods for sharing out these resources are called scheduling algorithms. The below paper sets up some experiments on the popular Xen hypervisor (this is what Amazon uses for its cloud offerings), and two of its popular scheduling algorithms. The researchers want to see if different combinations of CPU, network, and disk-usage in a system with two virtualized operating systems, would interfere with each other. The lack of interference is called Isolation. In general, Xen’s default schedulers were found to provide good isolation in high throughput, non real-time application environments. Once you move to real-time operations, you must modify the scheduler’s configurations to meet your needs.

The preceding is a summary of the research paper cited below.

Somani, G.; Chaudhary, S.; , “Application Performance Isolation in Virtualization,” Cloud Computing, 2009. CLOUD ’09. IEEE International Conference on , vol., no., pp.41-48, 21-25 Sept. 2009

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: