Welcome to Virtualization and Beyond
If You Want to Virtualize, You Only Got Two Choices
Written by Bill Baker, Sr. Product Marketing Manager, SolarWinds
Virtualization allows you to easily provision and grow
servers, as well as run more workloads on fewer servers. Essentially, it's "do
more with less." Virtualization does enable IT to run more workloads on fewer
servers, but it has become too easy to provision new virtual machines (VMs), and too hard to
manage them. This is causing customers to lose track of VMs that are no longer being
used, and consequently, the resources they are wasting.
In addition to wasted resources, some
organizations have discovered that as they deploy more and more VMs, their virtualized
environments become more difficult to manage, their application performance and
availability decreases, and their IT cost increases.
Virtualization offers distinct advantages like
improved consolidation. But every technology advance brings its
own unintended consequences. With virtualization, many data centers are facing uncontrolled
growth and over-allocation of resources to VMs.
What is VM sprawl?
VMs are easy to define and provision, and
users tend to over-strain their IT department with overwhelming numbers of VMs.
VM sprawl occurs when the
number of VMs on a network reaches a point where the virtual administrator can
no longer control the inefficient use of shared virtual resources (wasted or
misconfigured shared resources).
Why does it matter?
VM sprawl is a big problem that's costing companies time and
money. Too many virtual machines pose technical challenges and risks to the virtualized
infrastructure that can lead to infrastructure management issues, higher costs
due to unnecessary VMs, and reduced application performance caused by shared
resource contention.
Four Ways to Avoid the Pitfalls of VM Sprawl
- Get a Grip: Virtual administrators should
not waste time despairing that VM sprawl is out of control. Rather, they should
be figuring out why it is happening and how to make it manageable.
- Stop VM Sprawl
Before It Starts: The
best way to manage VM sprawl, of course, is to avoid creating it in
the first place. Identify and maintain an owner for every VM in the data
center-ensure that unattended servers don't become a liability.
- Get Rid of Unused
VMs: Frequently review
unused VMs and determine if they need to be permanently deleted. With plenty of
notice, start shutting unused VMs
down. If someone screams, you can power them back on. If no one screams, then
wait a few months, back up the data, and get rid of the VM (as well as any
snapshots and related disk files).
- Get Rid of Orphaned
Virtual Disks: Delete leftover virtual
disk files not linked to any existing VMs.
When it comes to enterprise computing, it seems like history often
repeats itself. First we had server sprawl, then virtualization-enabled server
consolidation, and now we have VM sprawl. If your IT is going to continue to virtualize,
then as Silvio of The Sopranos®
fame would say, "You only got two choices." Those two choices are either
investing in identifying and tracking the VM resources you need to align with
the demand, or risk degraded application
performance and costly expenditures for unnecessary VM resources.
Read more articles from the Virtualization
and Beyond Series.
##
About the
Author
With 30 years of IT industry experience across
multiple corporate environments, Bill Baker currently serves as Sr. Product
Marketing Manager for
hybrid IT performance management software provider SolarWinds, where he focuses specifically on product
marketing initiatives for the award-winning SolarWinds® Virtualization
Manager and SolarWinds Storage Resource Monitor solutions.