In this recent interview with Top Operator Journal, our VP of Sales, Arun Shenoy, talks about the problems with enterprises owning their data centers, the shift to the cloud, and the future management of data centers and IT management.
Top Operator Journal // July 2018
Farmers are hardworking people, but across the centuries they’ve steadily made life easier for themselves. Medieval times were a Back-Breaking Age for farmers thanks to never-ending manual tasks: tilling the soil, keeping plants watered and weeded, harvesting crops, and feeding the livestock.
In the Mechanized Age, farmers’ lives took a big turn for the better. A farmer could increase his crop yield a hundred fold by owning a truck, a tractor, and harvester machine. In turn, the farmer spent less time working the fields and more time in the garage as a mechanic.
Then, in the late 20th century, farmers entered the Outsourcing Age when farmers learned they could take away the expense and headaches of owning a harvester by renting one in October from a firm who specializes in loaning out and repairing farm machines.
Finally, in the 21st century, farmers entered the Virtual Management Age when a farm could be fully overseen by a managed service provider (MSP). Monitoring the status of his farm operation in a portal, the MSP recommends: which farmlands to own, what crops to plant, which processes to improve, and what equipment upgrades to invest in.
And these decisions can be made from the comfort of the farmer’s air-conditioned office. In turn, with his farm management hours freed up, the farmer had more time to market his crops and cultivate relationships with his distributors and grocer customers.
OK, I recognize this little history of farming is a bit of a fairy tale. Yet, there are strong parallels here to the future management of data centers and IT infrastructure.
As enterprises steadily move more infrastructure to the Cloud, the maturity level of their data center operations lies somewhere between the Mechanized Age and the Outsourcing Age.
But one particular data-center operator, Serverfarm, is taking its enterprise customers fully into the Virtual Management era. Through its deep, software-defining of data center processes it’s enabling its customers to get a detailed and consistent view of its infrastructure in several locations. And besides taking away the details of physical asset management, Serverfarm advises customers on all aspects of physical asset planning and operation.
This fascinating business model as explained in nice detail in my talk with London-based Arun Shenoy, Serverfarm’s global VP of sales and marketing.
Dan Baker, Editor, Top Operator: Arun, I know that Serverfarm has been a U.S. based data center operator for the last 17 years, but I understand you recently purchased your first European center in the U.K. Tell me about that.
Arun Shenoy: Yes, Dan. As our customers in the U.S. scale up, many seek to expand internationally and we’re taking them there. Our first investment in Europe kicked off in early March when we acquired a facility at Heathrow Airport. Then, in three months’ time we released much of the available capacity at the site to one of our U.S. customers.
This shows how our business works and how flexible and agile we are for our customers. To us, the data business is not constructing big shiny buildings. We select operationally capable existing sites then do a kind of sale/lease-back deal with the existing tenants so that new customers can come in and take up the available capacity.
That’s been our strategy for 20 years and we’ve become good at it. So today we’ve got our eye on several hotspots in Europe: Dublin, London, Frankfurt, and Amsterdam are the top four.
Interesting. So what market forces are enabling data center capacity to open up for you?
Well, about 15 to 20 years ago there was an explosion in enterprise-owned-and-operated data center capacity. So at the time, enterprises got highly competent in building/maintaining data centers of a relatively small-sized real estate and facility management environment.
Today, enterprise data centers own about 50% to 60% of the total capacity that’s been deployed. But the boom times have gone and many of these enterprise data centers are now vastly underutilized. Collectively they are probably running at only 30% of their capacity today.
There are many reasons why this happened. For one, enterprises saw a greater need to expand their capacity in new geographies, not at existing sites. Then, trends like virtualization — in compute, storage, and network — also changed the game.
Bottom line, most enterprises today figure it’s hard to create economies of scale in data center services as a core business.
Now it’s not that our customers have given up on data centers. All of them consider moving to the Cloud to be very important. It’s just that an even bigger issue for them is getting rid of the headaches of managing the physical aspect of data centers.
What are the consequences of enterprises getting out of the data center building and management business?
Dan, once they failed to invest in their data center development skills, they grew dependent on the infrastructure they built 15 or 20 years ago. Then, most of them outsourced the operation of those facilities to companies like CBRE and JLL. By doing that, they gradually lost their understanding of how the buildings, fabric, and all the physical aspects of a data center work together. So the enterprise no longer knows how to best evolve and modernize their infrastructure.
So as an enterprise, hiring a service provider like Serverfarm is pretty attractive because we can run their infrastructure regardless of where it is and how many pieces of infrastructure are in it. And we can manage it well from three key aspects: 1) an asset view, 2) a capacity view, and 3) future-change view.
With our approach, an enterprise can buy an outcome rather than a set of things. We offer them our real estate competence, our colocation competence, and our operational service competence wrapped in a platform we call InCommand.
Think of InCommand as a kind of service level agreement — an SLA that manages the assets they own, where they are located, and exactly how their assets come together and are used. Plus we manage all of the changes associated with those assets. The end result is infrastructure that is more reliable, more efficient, and cheaper to run than when they do it themselves.
So it sounds like a more comprehensive outsourcing service.
It is. The customer trusts us to take care of all of the physical assets we are putting together for them, not just the big things like generators and UPS, but everything in the data center: interconnect cables, switches, fiber connectivity racks, how distribution units move in data centers, I mean literally everything.
And that allows them to spin up infrastructure in the U.K. and in other European countries as we quickly move forward. But the key thing is they get absolute consistency across that experience whether they are with us in the U.S., East Coast, West Coast, in Canada, or in Europe.
Can you give me an example of what you mean by this “consistent experience”?
Yes, it means even though the infrastructure comes in a wide variety of buildings, equipment, and platforms, to the customer we make it look and feel like common Lego blocks.
We have eight data centers (a total of 100 megawatts across them): seven centers in the U.S. and one in the U.K. Yet no two of them came to us in the same way. Our Moses Lake facility in Washington State was an old U.S. Air Force command and control center. Our facility in Atlanta was a semiconductor manufacturing company. And our Santa Clara site came from another enterprise.
So, our skillset is to take very different facilities from different places and make the experience of operating them very consistent. And the way we did that was to stress three key things in our infrastructure management:
- Cultivate the best people — develop and invest in keeping their skills top notch;
- Make our process reliable and consistent regardless of a site’s age and equipment; and,
- Create a set of tools, a platform, that brings all those things together, so when our customers ask us to do something — regardless of where they ask us to do it — the outcome is what they expected — nothing’s lost in translation.
And I assume the customer accesses the platform in a portal.
Yes and the portal is part of the InCommand platform and Heathrow is where I think it gets really interesting. Lots of providers in the data center/colocation space make portals available to their customers. The difference here is that InCommand is also the platform we use to manage the infrastructure itself.
InCommand take us all the way from the very top in terms of how we engage with our customers right down into our own operation environment — how we manage our network operations center, manage our maintenance, alarming and reporting.
The InCommand is really the operating system for our data centers but I think even more importantly, our customers get the opportunity to engage with that operating system. And we make it look and feel like they are right in front of their facilities even though they may be deployed on another continent.
You talked about the importance enterprises put on moving to the Cloud. How do you help them achieve that?
“Moving to the cloud” is an enterprise goal, but it’s a moving target. One of the key challenges is properly dealing with the infrastructure that cannot move to the cloud right now.
Some infrastructure may never move to the cloud simply because it would be too costly in terms of re-architecting or re-developing. So the pain of managing physical things is not just about moving them to the cloud, but also managing the things that remain behind. Serverfarm steps up to address both sides of the equation.
Another service of vital importance is advising our clients on managing emerging capabilities such as 5G, IT, GDPR, and cloud. Every single one of those things will impact our customers’ infrastructure choices, so we become a key enabling partner for them in those spheres.
Arun, thanks for this briefing. I’m intrigued how you’ve wrapped a software-defined networking mindset around the business of managing data centers. It’s a powerful business model I think more and more enterprises will find attractive.
Thanks, Dan. The mark of succeeding in a business like ours is delivering the expected outcome. I think if you ask our customers, each of them would say we have delivered an infrastructure that is more reliable and more efficient than it was before we took them on as customers.
In many ways, delivering value on the real estate side is the easiest part. More importantly, our customers appreciate how we take away the pain of managing physical assets and create an evolutionary path as the world of data centers and cloud-based infrastructure constantly morphs into newer and better capabilities.
Click the link below to read this interview and other great articles from Top Operator Journal.
About Serverfarm
Serverfarm is an expert data center real estate and operations organization that delivers a high level of flexibility and personalization, ensuring clients receive solutions designed specifically to their individual needs now and into the future. Our mission is to maximize data center and IT infrastructure efficiencies by providing a complete approach to managing global data center portfolios.
