In good economic times or bad, it is always prudent to optimize
operational efficiency. One way to do this is to reduce costs. It
has usually been assumed that virtual server-style technology is
one investment that enables IT cost reductions.
Is this assumption true?
Well, yes...and no.
Gartner speaks to many clients about their server virtualization
investments. Historically, the proof points for return on these
investments have usually centered on reductions in CAPEX, that is,
reductions not only in regards to data center facilities but also
in hardware spend and associated maintenance. Physical server
proliferation tends to become controlled, and existing as well as
new server investments become more effective shared resources. So
far, so good. But what about OPEX?
OPEX usually includes things such as facilities (power and cooling)
as well as human labor costs (salaries, bonuses, insurance etc),
with labor expenses typically being the lion's share of any IT
budget. Energy savings may become one of the primary goals (and
hence justify the cost in terms of delivered benefits) of your
virtual server investment. Still, that leaves us with the labour
impact, which is critical if we are to derive a realistic ROI.
It is in the category of labor or ‘system administration
costs’, that the utility of the server virtualization
investment becomes somewhat less clear. Just as Amdahl's law says
that the speed of the ‘system’ is governed by the speed
of the slowest component, the benefit of the virtual server-related
investment will, we believe, ultimately be governed by the cost of
the largest associated component—systems administration. But
how to measure costs?
There are few models available to assess the impact of server
virtualization on system administration costs for IT organizations.
This is exacerbated by the fact that many organizations do not
baseline system administrative costs in a non-virtual environment
for comparative purposes. Organizations should develop a server
virtualization model to assess total system administrative costs
that includes activities such as root cause analysis, to account
for the (often soft-dollar) costs brought on by the added
complexity of a virtualized environment to obtain a realistic
summation of the overall benefits of this investment.
One option is to borrow approaches typically used to measure
software complexity. A software complexity-based approach using
both white box and black box techniques may offer a means to
establish an appropriate costing framework, irrespective of the
activity being analyzed. In the white box method, we need to
understand the internal structure of the program (lines of code,
number of ‘if’ statements etc).
With respect to our virtualization context, a substitute would be
to aggregate the timing of the administrative steps within a
service response to an end-user request. The use of a ‘white
box’ style methodology works when the specific steps are well
defined. Alternatively, use a ‘black box’ approach when
the specific administrative procedures are unknown or immature,
because if we can't see ‘inside’ a process, we may be
able to infer complexity (and thus cost) from an external
perspective. Using this method, we seek to equate cost as a product
of the number of interdependencies.
These ‘soft costs’ don’t necessarily weaken the
argument for investment in virtualization technology, but they
could reduce some of the net positive OPEX impact—the degree
to which this occurs would depend on the values assigned to the
variables in the models above. The key point, though, is to not
blithely assume only an upside to the operational impact of server
virtualization. In so doing, you will enhance your credibility with
the key decision makers in your organization and, through exposure
of the potential problems, lay the groundwork for improving these
potentially problematic processes.
Cameron Haight is a Research VP at Gartner Research