Abstract:

In a shared resource environment like Linux on z/VM, success depends a lot on how well you manage the use of the shared resources. Misbehaving virtual servers impact performance or require an excessive amount of CPUs to run the workload. This presentation shows how to measure CPU usage, understand where the CPU cycles go, and how to identify misbehaving applications. Where applicable, real life examples from ESALPS will be used to illustrate the process.

Rob van der Heij has been involved with VM since 1982, initially as a customer, later with IBM Global Services in The Netherlands. He has been working with Linux on z/VM in various roles since early 2000. Currently Rob is Software Engineer with Velocity Software GmbH in Germany developing the ESALPS Performance Monitor. He also does Linux on z/VM performance research and provides tuning support for customers. Rob is regular speaker at conferences and contributed to several IBM Redbooks. His professional interests include most that can be done on z/VM, in particular Linux and CMS Pipelines. Rob has been a very active plumber for more than 15 years.

	Close this Window
	Last Update: Monday, 02 November 2009 @ 22:53
	(c) Copyright 2009 GSE UK Region