Beyond SDDC

By : |July 7, 2015 0

MUMBAI, INDIA: In part 2 of the interview, Raghu Raghuram, EVP & General Manager, Software-Defined Data Center, VMware, uses customer references to clear the air around SDDC and emerging technology, network virtualization.

What was the genesis of software penetrating each and every segment of the IT infrastructure?
The traditional model of infrastructure quickly runs out of steam—where you have got your server and your storage and your network all being driven by your old style. Being able to accumulate enough of them to serve the volumes becomes impossible from a pure cost point of view and then being able to configure them and program them every day in order to deliver new applications becomes next to impossible and that was the genesis of those companies saying what if infrastructure could turn into software.

It turns out the way to do it is through virtualization. If you look at Amazon, Amazon is a cloud model built on top of server virtualization. But this is not sufficient. Therefore, you got to virtualize the storage and network.

___________________________________________________________________________________________________________

I think IBM was the first to speak about storage virtualization.
Yes, so IBM tried storage virtualization I would say in the early 2000s. There was a combination of factors at that time as to why there was not much pickup. Infrastructure is always driven by application need and the application need was not acute.

In early 2000s if you said, Hey I want to try my software changes and reprogram my storage every day! People would say, “Why do I need to do that?” There was not a need for scale and hypervisors had not become truly prevalent.

Until you solve server virtualization, storage did not matter. All that has changed over the last decade which is why now it is even more important, and ironically IBM has exited its storage virtualization business.

And network virtualization was entirely new that we did in-house and we also acquired a company called Nicira. The thing that we have done and that Google has not done is we have focused on the enterprise customer.

Google is a phenomenal company. They were building their own applications—Search, Gmail, Maps, etc are their own. They control the application and so the infrastructure can be tailored to their application end-to-end. Whereas, we serve the enterprise market and every enterprise is different-they have plethora of existing legacy applications plus they are building new applications, so how do you provide a one solution fits all application?

Does it have relevance to application economy?
Most certainly yes.
Let us look at the Bank of America, a 100 year-old company that has released a new application which now lets me deposit cheques and complete transactions from India compared to the past where I had to physically go to the bank for all transactions.

Other similar examples are United Airlines, Starbucks, E-Trade, Expedia, NFL Football, Mercury News, Wall Street Journal, and the newer companies like Uber, Amazon.

Secondly, consumerization of technology is the demand of the millennial generation, and enterprises are releasing applications the next gen will resonate with. So consumerization as a term has lost significance as the only applications that will actually get traction are consumer applications.

What then is the logical progression?
Clearly there is a whole set of and especially regulated industries that are looking to transform their internal data centers but everyone of them will tell you that they will use external clouds as well, and that is where the hybrid cloud concept came about. The key to enabling this world is having a common networking and management fabric and a common security model.

And different companies are in different stages of putting these things together. There are certain companies that say I am going to stop from the cloud and work my way back. There are certain companies that say I am going to start by modernizing what I have got internally and work my way forward.

It is the data requirements. It is regulatory constraints. It is how much networking has to happen back and forth. And it is the skill set of the people involved. Is the application highly bursty or not? What are the application characteristics?

Who is at the forefront? Where do you see the traction coming from?
We believe that the best way to enforce these regulatory environments and drive the new generation applications is cloud architecture. But these companies are deploying the cloud architecture for a large part in their internal data centers. They are experimenting with some external cloud models as well but they are not moving to the external cloud model any time soon.

They want to get educated on that and they want that environment to settle down. But they feel a lot more confident about their internal, transforming their internal architecture to a cloud model and in fact that is driving a lot of interest in our technologies inside this country.

The companies that are one in the cloud are small gaming companies or a Snapchat or a Pinterest-those companies that wanted a global audience right from the get-go even though they were like a 100 person company they have chosen to start from the cloud and then those that go in between. There are some applications that live externally and some applications that live internally and those tend to be more hybrid. So we see all three depending upon what your perspective is.

Does this mean paradigm shift at the back-end?
The set of technologies at the back-end do not change. But there is a need to scale-up. Previously an enterprise application could be a 9 to 5 application, but now with anywhere anytime access, all consumer facing applications are available 24×7.

Recently, in the US one of the most highly watched sporting events was a pay for TV boxing event-Manny Pacquiao and Floyd Mayweather. It was on a massive scale—the event would not have been possible without network virtualization technology because that was a technology that allowed for two things, one is automatic scaling and rapid deploying of the new application for this event.

I mean they put up a new Web portal that had the boxer’s histories and ordering and other things you could buy. And that was put together in a matter of months-to write the software but then the deployment of that happened almost instantaneously because of network virtualization and the entire scaling. The demand for people ordering the online paper view, purchasing and then watching the event, all of that was powered with the network virtualization technology. So this was a multibillion dollar revenue event for the casinos and that would not have happened without technology like network virtualization.

In India, there is a rapid requirement among banks that are setting up branches in remote locations. One example I want to give is the large e-commerce giant who announced a billion dollar sale. To accommodate the large volume, a very large bank had to rapidly open up payment gateways because a lot of purchases were happening that day and the e-commerce giant uses the bank’s payment gateway to handle the peak traffic.

Another instance is the hi-tech companies that serve global customers; these are isolated call centers or isolated development centers, which still operate largely on a physical model. But what has them super excited is a more virtual model they can literally create a virtual data center by using software infrastructure. Just think about the business as well as the infrastructure value of that!

Banking is yet another great example where bankers want to put-out consumer facing application but worry about security and so they say, look I want to deploy network virtualization tomorrow morning because I have got to deploy my application to the front-end.

An interesting trend that has come-up in most of these enterprises is that the line of business (LoB) heads, in this case the digital marketing officer, who normally own the event/s are overriding the infrastructure teams.

Agreed. Self-service driven LOBs are driving investments and two there is a Board (including the CIO) which is taking fundamental decisions. Many CIOs are worried …
Yes, absolutely. Different companies in different industry verticals are approaching it differently. But the digital officer has become in many cases a pretty key executive officer of the company. In some cases, it is a role that grows with the digital marketing… I mean with the CMO and the CMOs are transforming into the Digital Marketing Officers but in many cases it is going well beyond that because the entire business model of the company is now a digitized business model.

Those are the places where Boards and executive teams are thinking about this in a very strategic way. There are other places where that has not fully happened yet and the CIO is driven by pressure to be more relevant to his business peers or driven by a need to dramatically change the economics of their IT model are adopting these sort of technologies. So those are the high level business drivers that we see. And that applies universally, it does not matter which country or…

This movement is becoming so pervasive that wherever you go, your ability to get your next job depends upon showing digital success. Nobody wants to hire a CIO whose resume says, Hey, I ran a hardware-based data center, people are just not interested in that. Not only are you going to get fired from your current job but you are not going to get a new job. This six month, 12 month imperative now becomes-in 12 months I want to show digital transformation or some that I have initiated that or executed. If anything, this is putting more pressure on the CIOs because the CIOs that are going to be successful are those that are going to be at the forefront of how paying organizations make the shift.

And that is the cream that is going to rise to the top. The average tenures are not changing dramatically. But the need for them to show results quickly exists and that – because now we are living in a reputation-based economy. Especially, in a loosely knit society like CIOs, you have got to be seen as someone that fundamentally pushed your company into a software-based world.

What is your advice CIOs who are currently penning new strategies for their enterprise?
Fundamentally the question is where does the business want to go next? And CIOs have two opportunities there.

One is actually educate their peers on what a software-based model for the enterprise can achieve for them in terms of making companies more digital. And second is providing a holistic execution path and that enables them to achieve that. It has to start from the CIO becoming a partner to the business, partner to the digital marketing officer of the digital, whatever be the title is and driving from there on down.

I would say that should be what the forward looking CIO should be thinking about. What are the new business models that a CIO can now enable for their business. The second one of course is the transformation and the preparation for that. And that would say, how will I make my existing infrastructure more efficient, how do I release dollars for new application deployment? And both those could be starting points for what CIOs think about.

How can companies like VMware help?
We help at two levels. One is like I said before, SDDC is a fundamental enabler of the cloud and we help these companies build the cloud.

The second area where we help is on the endpoint. AirWatch is our technology for managing and securing the endpoint. And today the bulk of the endpoints are these types of devices but we have been working with some of the early adopters and technology providers to make the same mechanism work on IoT and points as well. We provide the fundamentals of the technology enablers whether it is a smart city project or a very specific IoT for automated machine floors or monitoring of industrial equipment and so forth. That is one way by which we play in this market.

The second is we bring about the business education to help businesses understand what other similar businesses are doing around the world. So, in many ways we have started to advise CIOs on how they should think about their infrastructure of the future to serve the needs of big data and serve the needs of IOT, serve the needs of mobile and the digital end user experience.

Also read: Demystifying SDDC

No Comments so fars

Jump into a conversation

No Comments Yet!

You can be the one to start a conversation.