Subscribe


White Papers

Free IDC Vendor Spotlight

Converged Infrastructue And Today's Buisness Needs

download now
Demo & Video
Untitled Document

Introduction to Converged Infrastructure Video

Product Overview

Is Disaster Recovery the next killer app for virtualization?
Posted March,11,2008 by Pete Manca
There have been a slew of articles speculating that Disaster Recovery is the next killer app for Server Virtualization. The premise is that since the OS and application are abstracted from the hardware, it’s simple to move the entire stack from one location to another, and presto – DR solved! Unfortunately, it’s not that simple. While Server Virtualization goes a long way towards helping DR, there is still the issue of the physical machine management. Hypervisors require similar if not exact hardware configurations on both the local and remote systems. And applications that require DR typically use external storage and complicated network topologies. These too have to be configured identically. All this means that there is more to DR than moving a virtual machine from one server to another. To truly solve the DR problem in a simple, cost effective manner, you need to solve all levels of complexity at the same time. Virtual machines take care of OS and application abstraction layers, making those components easy to move. Add in a Virtual I/O infrastructure that can virtualize the storage and networks, and you're almost there. The remaining piece, and the one that is most often overlooked, is an Infrastructure Virtualization or abstraction layer that can represent the data center components – both physical and virtual - in a logical manner and in a way that can be replicated from site to site. Remember, applications are not always 100% virtualized. Think of a 3-tier app where the web server and app server are virtualized but the database server remains physical (usually for I/O performance reasons). An Infrastructure Virtualization layer can abstract both physical and virtual servers, and allow for simple, low cost DR by replicating that environment on remote servers. It also serves as a hardware abstraction layer so dependencies on server configurations are removed. So, before anyone buys in to the panacea that server virtualization solves DR, please look at the details and carefully understand what it can and can’t do. Then maybe you’ll see the value (PANecia…) of what an Infrastructure Virtualization solution like PAN can do.

Leave a Reply