Monitoring performance in virtualized environments
A recent InformationWeek piece highlighted the real elephant in the virtualization room, and that is the question of performance. Abandoning the model of running applications on a dedicated server, and provisioning (or more likely, overprovisioning) storage space to each application, in favor of a virtualized environment has its advantages, and there's no doubt that it's the future of computing. Virtualization saves an enormous amount of time and money. It's a game-changer. But guaranteeing the performance of applications running on a virtualized environment is not quite as straightforward as doing so in a dedicated environment.
The question of performance is a common objection, but it can be addressed effectively and need not serve as an excuse for resisting the inevitable march towards virtualization. In many ways, a server running a virtualized environment isn't that much different from a server running a dedicated application. Monitoring tools, both for monitoring the applications themselves as well as the server's CPU utilization or the storage device's storage utilization, are available to make sure IT personnel receive alerts in case of a problem. Such monitoring isn't always direct though, with applications moving between servers, but systems management tools are rapidly stepping up to the plate with new innovations to apply traditional systems management tools to virtualized environments, and this added visibility and "logical map" will go a long way towards removing the objection of performance and performance management from the virtualization equation.