CloudSwitch Aims to Solve the Risk and Complexity of Cloud Computing

Ellen Rubin (ER): In general, we feel most people connected to the IT world understand the benefits of cloud computing. Certainly, the general theory behind it makes sense: if you can gain flexibility and drive down costs by using someone else’s infrastructure to run your applications, you should do it. But, reality is a little trickier.

There are four major things that need to be addressed: security, integration between the datacenter and the cloud, simplicity in mapping the application into the cloud, and removing the issue of vendor lock-in. The general premise of CloudSwitch is to solve the risk and complexity of moving enterprise applications to the cloud without requiring—and I stress this—a huge engineering effort to get there. Using CloudSwitch, the user doesn’t have to modify the application at all to fit the specifications of the cloud provider.

JP: So, you have a solution that addresses some of the key hurdles we hear about time and again, like security and application portability, without over-stressing the IT team. Can you talk a bit about the solution itself?

ER: Absolutely. First, it’s a software appliance that gets downloaded as a virtual machine inside the enterprise datacenter. The appliance contains the management components for discovering applications, orchestrating cloud deployments and managing cloud utilization. Our Cloud Isolation Technology creates a secure, encrypted tunnel that bridges traffic between the datacenter and the cloud provider. It deploys CloudSwitch instances in the cloud that manage the “local” environment based on policies set by the customer, while integrating into existing datacenter tools and processes. Through the Cloud Isolation Technology, we are able to abstract some unbelievably complicated stuff from the user. This is our core IP. It’s an abstraction layer that does the automation of the mapping between the on-premise application and the cloud provider. And by application, I mean the entire application stack, including the database, OS, and networking configuration. So the customer can move applications with point and click simplicity between the data center and the cloud and back.

JP: And if you want to know more, you better sign an NDA, right?

ER: Pretty much. I mean we are more than happy to get on a whiteboard and get into the technical details but, as I said, this is our core IP.

A major point, though, is that CloudSwitch enables applications and data to be secure at all times—whether in transit, at rest, or during processing. Our technology protects the communications between the data center and the cloud, all network activity in the cloud, and all storage used by the applications. The Cloud Isolation Technology does all this in a highly automated way.

As a result, companies are no longer bound to a particular cloud provider, the enterprise gets total cloud independence and the ability to choose the cloud provider that best fits its needs, and the user can bring the applications back to the enterprise whenever the they choose or conditions warrant it. And it’s all done in a secure way with continuous control.

JP: And this isn’t just theory, right? You’re just about to finish your initial private beta. How many customers were involved? What have you learned from them?

ER: Well, our private paid beta, which is coming to a close this month, consists of six customers: one each from the healthcare, telco service provider, clean tech, drug test/discovery industry and two from the pharmaceutical industry. They were either large or medium-sized companies.

In each case, the customers had some experience with virtualization and they had a compelling reason to experiment with cloud computing. To get started, these customers moved back-office apps, not mission-critical or high transaction apps. They wanted to take an application and move it out to the cloud; test the user experience; get a feel for the latency. And, most importantly get a feel for what it’s like to manage an app that’s running on a cloud provider’s infrastructure.

We learned a ton of stuff.

For example, we realized it was very important to make it very easy to install run. We realized the need to make the network configuration process easy, so that getting out to the cloud was painless. It was clear that if it turned into an extended engineering project, we’d be in trouble.

Using CloudSwitch does not mean a new project for the engineering or networking teams. It ties into current processes. That’s big.

So, we worked really hard to improve the user experience and to remove as much complexity from the experience as possible. We’ve made it intuitive to install and run, so customers can get into the cloud quickly.

We also learned that enterprises are in the beginning of understanding the cost effectiveness of cloud computing and which applications are most effective to move. And, now that we’ve worked with live customers it’s been very interesting to see the real proof. It’s an IT decision with all the same dynamics that any IT decision encounters.

JP: So, looking down the road, what’s in store CloudSwitch?

ER: Near term, we’re focused on making enterprises’ initial cloud experiences be as successful as possible. And we’re making sure that our product is integrated into the existing tools and processes within the enterprise data center.

Our roadmap has a lot to do with adding additional cloud providers and building in additional functionality to make the product as widely available as possible. Beyond our first target platform, which is Amazon EC2, we are working and listening to customers to prioritize those that will be available next. We’ll have more to say about this soon.

Longer term, our founding vision is cloud brokering. In other words, after enabling enterprises to connect their datacenter to the cloud, the next goal will be to take advantage of a heterogeneous world of cloud-based services. Cloud brokering will be based on business and technical decisions. For example, decision criteria based on SLAs, latency, cost, business rules, and regulatory rules. So, across different clouds you’d be able to pick the one you prefer based on these criteria.

JP: Let’s come back to the things that are holding cloud computing back from wider adoption. What issue is top-most in your mind and why?

Well, as we’ve discussed the main issues holding people back—security, integration, lock-in—are well understood. We may not agree on how to technically solve them, but we know what they are.

But, everyone’s in the early days of this transformation and people are trying to figure out how exactly to do it. Some are considering a private cloud or some hybrid approach using a public cloud. And people are trying to map out the relationship between them. So there’s a lack of knowledge and education, combined with lots of experimenting. All the stakeholders need to be brought in and assured this thing won’t blow up if they adopt it. And, yet, interest is only growing. So, I tend to think we’re reaching a more pragmatic stage. That is, where companies want to get more serious about cloud computing and start to get realistic about how it will be used in the enterprise. I think many people are coming to the conclusion that it’s primarily the most sophisticated operations that have the expertise to build their own cloud environment.

JP: It almost sounds like we’re reaching a stage where cloud computing becomes a build vs. buy proposition. What do you think?

ER: Yes, I think that’s right to a large degree. At this point, many business and IT leaders have learned enough about cloud computing to know they need to pay attention to it. They also realize the big roadblocks, the things we’ve discussed. And, so, they have some idea of what is involved to move to the cloud. Because enterprise IT is about legacy systems, control, and internal politics. And they’re stretched thin, so experimentation is a challenge. So they’d like to be able to use the best aspects of the cloud in the most pragmatic and cost-effective way possible.

JP: Well, that brings up an interesting topic. What are your thoughts on the public vs. private cloud debate?

ER: Enterprises are turning to both kinds of clouds to reduce costs and improve responsiveness. You can learn from a lot of the stuff Amazon has done, but they have built this over many years. That’s not going to be easy to replicate. So, you have to ask yourself, how do you make the enterprise more cloud-like?

Most companies will want to take advantage of public clouds. Think of it as a tiered approach. I’d say about 50% of a company’s apps could potentially be moved to the public cloud. There’s no reason to own and maintain the infrastructure required to support things like test and development, back office apps, web front ends, or business continuity and disaster recovery. All of these areas are good candidates for the cloud.

But, large and medium-sized companies have lots of legacy stuff to deal with. And, it’s not unreasonable to hear them talk about private clouds as a “safer,” more controlled option. Maybe it doesn’t have all the same benefits as a public cloud, but it gives them a bridge.

So, we’re actually huge believers in hybrid approach: keep those mission-critical and/or “sensitive” apps in-house, but move whatever can reasonably moved out to a public cloud.

Of course, this topic will get a lot more heated and interesting especially with the release of Microsoft’s Azure platform. We are clearly headed into a much more heterogeneous world. Ultimately this is good for CloudSwitch—more clouds means a better opportunity for us to help customers move to the cloud and between cloud providers—at their discretion.

JP: In that case, how do you see the cloud computing market evolving? What’s your take on cloud federation and the concept of the intercloud?

ER: Well, this is another one of those debates that gets people arguing over what’s needed to propel cloud computing further. The argument for the intercloud is that standards are needed before you’ll see wide enterprise adoption of cloud-based services. Without standards customers won’t embrace it. But standards emerge over time.

We see things a little differently. While, we agree that the connection between clouds has got to get better, we think there’s enough happening today to get started with cloud computing. In fact, federation is one of the prerequisites toward the goal of interoperability and it’s already happening today.

JP: And that brings us back to the fundamental principal behind CloudSwitch?

ER: Yes, we think it does. CloudSwitch is an enabler of cloud computing for enterprise-class customers. We can help customers move applications to a cloud provider with minimal effort, while maintaining control and security over those applications. For us it is rocket science, but for users if you can point and click, you can do cloud computing.