Some good points made by Steve Wexler, especially with respect to Hardware. Should Hardware meets its demise, what is going to run the Software and what will the Software control? “Hardware Still Matters (Sort of. Maybe.)”
There are 3 major drivers to the adoption of Software Defined Networks:
2. Lower OpEx
3. Lower CapEx
Yes, these are in order of priority. The main reason that corporate IT departments have lost control of the IT spend is that it’s so easy (agile) for anyone with an expense account in Marketing (or any other business unit) to set up a virtual environment in the Public Cloud. It takes approximately a minute and can be as low as $5 per month… an amount that nobody questions on the expense account.
But what about the security of corporate assets? If IT can provide users an elastic and flexible service internally, they will solve the first problem… agility for the dev groups in the business units in a secure environment.
What about CapEx?
With SDN, you simply need less hardware and what you do need will be cheaper and faster. However, since CapEx usually only accounts for 10-15% of the Total Cost of Ownership for the IT infrastructure, you will have to make a big dent in this cost to materially affect the overall TCO. Gartner maintains that the hardware spend will actually move from 12% to 13% of the TCO with the advent of SDN.
The final issue is OpEx. This is where we need to spend out time. The automation of the orchestration of network services promises to relieve network operations staff of much of the manual grunt work associated with service deployment freeing them up to work on projects with higher value. Since the labour component is usually the largest contributor of OpEx in the TCO model, any reduction here should provide the material impact that CIO’s are looking for in the Network TCO model.
So why don’t more corporations have Private and Hybrid Cloud strategies based on SDN?