Virtualization and Cloud executives share their predictions for 2013. Read them in this VMblog.com series exclusive.
Contributed article by Rex McMillan, senior product manager at LANDesk Software
As Storage Impact Continues to Grow, More IT Departments Will Turn to Non-Persistent VDI
Currently IT administrators have two options when
implementing a VDI solution for their users. They can build a persistent image
for each user that is preserved when the user logs out. This is very similar to
a standard physical device, and the management agent can be "baked-in" to the
image. In this model there are minimal complications as the agent only needs to
be deployed at the time the virtual machine is initially created.
The second approach is to deliver a non-persistent instance
that is created each time a user logs in. The complication in this model is
that the management agent is not associated with the user, and each time a user
logs in, a new inventory record is created, resulting in multiple inventory
records that are obsolete. That being said, there are advantages with this
method as well - namely that with networked storage as their data store, many
users can leverage the same virtual desktop image file, significantly reducing
virtualized image storage requirements.
The growing need for storage space is a rather hot topic these
to Aberdeen Group, many companies will have to double the volume of
their data storage every 2.5 years just to keep up with storage demands.
Even with the lower costs of today's storage, investing in this
area is not a good option for many cash-strapped IT departments. One option they will consider to save space in
2013 is to investigate non-persistent VDI. While this may not seem like an
earth-shattering prediction to many, it is a worthy consideration across the
board for task-oriented or casual end-users whose desktop tool requirements and
processing roles are consistent. The only caution is... this method does create a
different computing experience and for knowledge workers in today's empowered
end-user environment, you can expect that this could cause some confusion and
In the non-persistent environment, all system changes are
ignored. Even though the machine operates as a normal desktop, as soon as it is
rebooted, all changes are lost. The loss of changes upon reboot means that
users cannot customize a non-persistent VDI as their "own" desktop. Considering
that many end-users like to create their own backgrounds,
screen savers and other customized settings, some may be frustrated by a
non-persistent environment. While this can be annoying for the end-user, it can
provide a benefit to IT as it could streamline operational efficiencies and
that if rogue applications are downloaded by end-users, they do not remain on
the system, where they could potentially expose the corporate network to various
security issues and malware.
If your IT department plans on making the switch from
persistent to non-persistent next year, my advice would be to make sure your
end-users know this change is coming. In my experience, folks are more
receptive to change when they know it's coming and don't feel blindsided by it.
Send out an email giving employees a head's up that this change will take
place; or better yet, set up a workshop where you explain the difference
between persistent and non-persistent VDI and how this will impact the
end-computing experience. You'll end up making both audiences you serve happy -
IT will save on cost and end-users will feel informed. It's a win-win worthy of
your consideration as you investigate choices for delivering better control and
About the Author
Rex McMillan is a
senior product manager at LANDesk Software, where he is responsible for the
well-recognized and widely-acclaimed Management Suite product line. Prior to
this role, he was a global systems support architect for LANDesk, where he was
responsible for designing customer solutions and resolving customer issues. He
is based in Salt Lake City, Utah.