UPDATED 17:00 EDT / APRIL 23 2013

Is AWS Eating the Enterprise? Not so Fast, Others are at the Table, Too

AWS is growing at a phenomenal rate. Wikibon estimates it will top $3B in 2013. But, says Jason Mendenhall, EVP for Cloud at Switch Communications, creator of Nap 7, a.k.a. SuperNap, the world’s largest, fastest public cloud colocation data center, it is not eating up the IT-as-a-Service marketplace — in fact in some ways it is helping that market by creating publicity. Nor is it going to eat up the internal ITO, at least not by itself, although it is helping to force major change in the organization.

Speaking at the Wikibon Peer Incite April 23 titled, “Software is Eating the World–Will Amazon Eat the Enterprise“, Mendenthall argued that no one cloud service, even one as large and innovative as AWS, can meet all the needs of all the different industries, companies, and government industries out there (full video below).

Several competitors, ranging from Rackspace to startups like Cloud Sigma, some of them running in the SuperNap data center, are prospering by offering services that AWS does not. Some, for instance, focus on meeting the specific needs of verticals like healthcare, with its special compliance requirements. And while some companies have a strategy of moving everything to the Cloud, others, again including some running on SuperNap, are keeping their critical systems in their own control. “People are no longer asking why cloud,” Mendenhall said. “They’re asking which cloud.”

These AWS competitors are growing right along side AWS because they provide advantages in one or more of four areas: performance, support, cost, and control. One major issue companies, particularly in highly regulated industries, have with AWS is the complete loss if control over their applications and data, which often creates problems with compliance audits. They simply do not know where their data is or whether Amazon’s infrastructure and operational practices meet the strict requirements of regulations like the U.S. HIPAA. For instance, Mendenhall said, the common practice of IaaS providers is to stripe client data across multiple arrays. What happens to that data when one of this arrays fails and is replaced? Is the sensitive patient data left on those disks, where it can be recovered after it leaves the AWS data center? That would be a clear compliance violation.

AWS and most other cloud providers co-locate data from multiple clients on a single system to save costs. Again many healthcare providers consider that a violation of security and patient privacy requirements. For that reason, companies in highly regulated industries either retain control of their core systems or at most move them to a service provider that specializes in their vertical and has the certifications to demonstrate that its operational environment meets or exceeds all relevant compliance requirements.

Big Data Logic

Even those who elect to keep possession of their core applications and data rather than trusting them to a cloud provider, however, are moving systems into very large, cloud-architecture data centers like SuperNap in part to get their valuable internal data into a facility that can provide access to public Big Data sources, whether those are social media, market data, or machine-generated logs, and to the large amounts of compute resource needed to analyze the combination of internal and external Big Data.

“The money is in taking advantage of their own data,” Mendenhall said. “But for most companies, building the high-performance compute environment that they need to do the analysis on petabyes of data is a problem. So they move the analysis part out to the cloud.”

But moving large amounts of data at Internet speeds presents its own major problems. “As people get into this, they discover that the network is the forgotten critical piece,” he said. “A 2 petabyte database is not that unusual, but moving 2 Pbytes over a 10 G line takes 10 days optimally.”

To improve that, Switch has pooled multiple fiber-based data network providers to offer speeds of 40G to 100G. By pooling the buying power of all its customers it gets these transmission rates at a fraction of the cost that even a large multinational company or government agency can command. It has been creating platforms & models based on different industries to use that capacity combined with the huge processing and storage power of SuperNap.

One reason that a large number of early Big Data adapters, including many of the world’s largest companies, have moved significant parts of their IT infrastructure to SuperNap is that it gets their data into the same room with some of the Big Data sources they want to combine with their data and with the compute resource they need to do the analysis. And the logic then is to move the applications generating that internal data — including both traditional applications like ERP as well as Web 2.0 & 3.0 applications built on cloud technology — into the same data center to obviate the need to maintain two copies of the data in separate locations.

The Two Aspects of Cloud

Cloud really is two things, Mendenhall says, that are impacting ITOs in different ways. First it is a new option for moving traditional applications out of the internal data center to decrease costs. This is the same outsourcing that IT has done for decades.

The second, however, is much more profound and impacts dev/ops. Both software vendors and internal dev/ops are adopting the cloud technology platform as the basis of new development to gain the core advantages of cloud flexibility, including, critically, the ability for applications to automatically expand and contract the amount of infrastructure capacity they use to meet changes in demand. This is particularly important for global enterprises that are likely to see sometimes unpredictable peak loads on specific applications and infrastructure. Whether these applications end up running on AWS or a competing IaaS provider or on an internal cloud, they are Cloud 2.0 or 3.0 applications because of their structure, which differs from traditional software in significant ways, and the open source technology platform they are built on.

Organizational Impact

These and related pressures ranging from hyper-scale infrastructure to mobile computing are combining to break the traditional siloed IT organizational structure and replace it with a service-based structure. If ITOs are to avoid being eaten by public cloud services like AWS and Salesforce.com they absolutely have to make this transition. And this means much more than defining the business services the ITO provides and reorganizing the IT budget to allocate infrastructure costs to this services. It must reorganize completely around them to focus on deliver the QoS that the business needs.

On the one side the increasing automation of underlying infrastructure, and in particular software-led infrastructure that unifies management of what is left in-house in a single pane of glass, along with the increasing incorporation of  cloud and outsourcing services like IaaS and Switch that moves the infrastructure out of the data center, are removing the need to maintain the complex lower level skills in manual hardware operations that mandated those technology based silos in the first place.

On the other hand, competition from the cloud and SaaS in particular is mandating the change. IT has to realize that business management has always seen IT in terms of the services — financials, e-mail, business management tools, etc. — that it provides. They do not care about all the technical issues that IT finds fascinating, they just want their BI to deliver the answer they need when they need it. Beyond that their only real concern is to be sure that the IT geeks aren’t running the computer equivalent of a huge and expensive model train set on company money.

So when Salesforce.com began to invade corporate C-suites to talk to the head of chief sales & marketing officer and CFO about CRM as a service delivered over the Internet at a specified usage based price and QoS, they understood it exactly. It was the utility model, and they embraced it even if internal IT could provide the same services more cheaply.

The lesson is that if ITOs don’t want to be eaten up by AWS and the rest of the cloud service providers they need to change in basic ways. Otherwise they will fight a losing rear-guard battle against more raiders from the cloud who go directly to business management with their offer of packaged business services, and CiOs will only know what piece of core IT will be next to go when the CFO calls to tell them about the latest cloud contract.


A message from John Furrier, co-founder of SiliconANGLE:

Your vote of support is important to us and it helps us keep the content FREE.

One click below supports our mission to provide free, deep, and relevant content.  

Join our community on YouTube

Join the community that includes more than 15,000 #CubeAlumni experts, including Amazon.com CEO Andy Jassy, Dell Technologies founder and CEO Michael Dell, Intel CEO Pat Gelsinger, and many more luminaries and experts.

“TheCUBE is an important partner to the industry. You guys really are a part of our events and we really appreciate you coming and I know people appreciate the content you create as well” – Andy Jassy

THANK YOU