What a difference a decade makes! About ten years ago when I proposed doing a storage virtualization pilot in my company’s data center, nearly everyone thought I was crazy. In the words of one person I was “introducing a production hydrogen bomb into our environment.” Contrast those thoughts with the conversation that I had late yesterday afternoon with the CEO of a data protection software provider who confided in me, “2011 will be the year that virtualization goes from early stage testing and pilots to full implementation across data centers of all sizes.”
In the 2001 time frame when I was tasked with managing my company’s storage infrastructure, the problems in front of me seemed almost overwhelming. While I was managing a SAN that was relatively large at the time (11 TBs of storage using two HDS 7700e’s that was servicing about 30 AIX and Sun servers,) it was manageable. It was what was not connected to the SAN and coming down the pike that concerned me.
In our data center (my company had multiple) there was over 300 Linux and Windows servers, about 150 AIX servers and another 100+ Sun servers that were still using local storage. Our rule of thumb at that time was due to cost constraints we would only connect those application servers that required more than 50 GBs of storage to the SAN. However I could see the day coming in the not too distant future where the cost of networked storage (NAS or SAN) would decrease and my company would issue a directive that all of those servers with local storage would need to be connected to some type of networked storage solution.
It was those challenges that led me to start looking outside of the box – literally and figuratively – for storage solutions that met my company’s needs. This eventually led me to identify storage virtualization software like that from DataCore Software and FalconStor Software. They had the right framework, at least conceptually, to take my company from the traditional “server-to-storage” connection to a more holistic, flexible solution where the applications and data could decoupled from the hardware on which they resided and freely moved to the right hardware platform.
This seemed to make perfect sense to me in mid-summer 2001 when I proposed heading down that path. Little did I know that it would send me on the journey on which I am now. At the time, I didn’t realize how much I didn’t know about storage and what obstacles – real and imaginery – that I personally and the industry as a whole would encounter and have to overcome in order to get to where someone can confidently tell me that 2011 would be the year that virtualization gets implemented.
Granted, in 2001, almost no one was broadly talking about server virtualization. Virtualization was almost always discussed in the context of storage virtualization. Then once server virtualization did start to take off, I did not see it driving the broader adoption of storage virtualization. If anything, once server virtualization adoption started I thought it might actually minimize or even negate the need for storage virtualization.
Instead, it appears to be doing just the opposite. As organizations recognize the cost-savings associated with server virtualization, they are looking to extend those same benefits to their storage infrastructure. They now see that with virtualization comes the opportunity to more efficiently and effectively use the storage capacity that they already own and will be purchasing and, in the process, derive the maximum value for their capital investment.
But maybe more noteworthy, and the reason for my conversation with the CEO of this data protection software provider, is that companies are awakening to the new opportunities that virtualization affords them to solve other long standing IT problems in the data center. The elimination of backup windows, doing near-real time failovers of their production applications to either other data centers or even to the cloud, and providing better methods for testing upgrades to their applications by using near real-time copies of data before those upgrades go live in production.
This is why I am so excited about this upcoming decade . I recently read that many CEOs and CIOs consider the last ten years to be a lost decade from a technology perspective. I couldn’t agree more, at least in this respect. Vendors and companies have been trying to make server and storage configurations work that should have been put to be a bed five if not ten years ago to make room for architectures that are better suited for today’s requirements. It looks that will finally begin to happen in 2011 and, in so doing, bring technology back in the business conversation again.