Recently I had the pleasure to have an extended and insightful conversation with Michael Crandell, CEO and Founder at RightScale, Inc. We discussed a range of topics related to how he sees cloud being utilized at the SMB and enterprise level, his perspectives on the industry at large, and how he sees the space developing.
Gathering Clouds: Looking out at the cloud industry, how do you understand it, how has it progressed, and looking forward, where do you see the innovation happening?
Michael Crandell: RightScale was born just after the advent of infrastructure-as-a-service clouds, namely Amazon Web Services with EC2 and S3. Our CTO and Co-Founder, Thorsten von Eicken, had been a colleague of Amazon’s CTO, Werner Vogels, from their days on the faculty at Cornell University, and when Thorsten found out about EC2 and thought about its potential, he got very excited. That was the summer of 2006, and Thorsten set to work building what became RightScale. Shortly thereafter, as Rafael Saavedra (our VP Engineering) and I joined Thorsten to start RightScale, we formulated several core tenets of our vision. One was that cloud infrastructure – with its API-driven provisioning, pay-as-you-go business model, and capability for automation – was going to be a huge game changer. It sounds like I’m stating the obvious now, but back then we were among the few initial voices claiming that cloud computing was a disruptive force and would fundamentally change the way IT is delivered and consumed. Our second core belief was that if that were true – if cloud really was that important – then it would spread and become a fundamental approach to designing infrastructure across many providers in the industry.
Looking back now, that vision has basically come true. If you fast forward to today, you see almost every large tech company on the planet has released an infrastructure-as-a service offering, the latest being Microsoft and Google, both in June of this year. So one key observation on the state of the cloud in the second half of 2012 is that infrastructure-as-a-service has really become the dominant form of cloud below the SaaS layer. Of course there are also many platform-as-a-service (PaaS) offerings, and certainly they’re important, and will continue to grow. They’re just going to take longer. For now, infrastructure-as-a-service has turned out to be the right mix of automation and standardization, but with customizability and flexibility. That’s why it’s gained so much momentum, and finally seems to be “crossing the chasm” in terms of adoption.
GC: You pointed out infrastructure-as-a-service as where things have moved towards. Where do you expect them to move to next, or where would you like to see them move to next?
MC: As we move forward we see that the lines are blurring between platform-as-a-service as it was originally introduced and infrastructure-as-a-service. Originally platform-as-a-service was a black box: you point the system at your code repository and say “here’s my JAVA (war) file, go run it for me,” and everything under the covers is handled automatically. The PaaS service scales the database, scales the app and web servers, and you don’t need to worry about any of that – that was the classic original platform-as-a-service. And originally the IaaS services were focused mainly on compute, storage, and networking resources – low level stuff. Today, however, infrastructure service providers are adding more discrete services, to the point where AWS now must be close to two dozen different web services ranging from things like RDS, database-as-a-service, email-as-a-service, Hadoop-as-a-service, load balancing-as-a-service, Elastic Beanstalk, and so on. So, more and more granular but powerful services are being offered that lie somewhere in-between the classic storage, compute and networking of IaaS and the full just-point-me-to-your-code-repository of PaaS. We’re presented with a quiver of ever more powerful and abundant services that companies and developers can use to assemble into powerful end systems. The opportunity for any developer or company moving forward is to figure out how to take advantage of these increasingly powerful services in the right mixture with basic compute, network and storage resources, and at the same time maintain flexibility across all of the cloud infrastructure resource pools that they want to use – whether they’re public or private, on-premise or hosted.
GC: So you touched on this a little bit with Microsoft and Google – but with their increasing noise in the space, do they actually present any real challenge to Amazon’s dominance? Can they really make a dent in what Amazon has been able to do?
MC: They will absolutely be players in the space. The way I’d state it is not necessarily that they’ll make a dent in Amazon’s business, because we think AWS is fantastic and that it’s going to continue to grow very fast. We see it more as the whole pie is going to increase and accelerate even faster with the addition of other major players. So yes, absolutely, there will be many players in the space. They will have to differentiate to build their own share of the market. No one company is going to provide everything that you need. Companies are increasingly sensitive about getting locked into an individual provider – what if the prices aren’t competitive? what if I need resources in a location that they don’t provide? what if I need regulatory compliance that they don’t provide? So, for all those reasons, companies really need choice, and today, more than ever, they have a plethora of choices. RightScale now supports nine different public clouds and three different private cloud technologies. And we see more and more companies using a mixture of those. In fact, a recent survey we conducted showed that 68 percent of companies said they were planning to use multiple clouds.
GC: Pulling from things you’ve touched on in terms of how RightScale is seeing customers use the cloud, in your view, what is the value for the variations of cloud that a small/medium business can use versus what an enterprise can use?
MC: That’s a great question. There are two ways to talk about the value of cloud. To start out from a global perspective, cloud has enabled a new generation of startup projects because of its easy consumption, its pay-as-you-go nature and its flexibility. These startup projects might take the form of individual startup companies, or they might be projects within larger organizations. The number of projects that we see being enabled by cloud computing is unprecedented, and many probably never would have had a chance to get started in the past. We view that enablement as a major social benefit in the business world. The second major value of cloud has to do with increased focus. Companies can now focus on their core differentiation –the secret sauce to their business – rather than on non-core, infrastructure concerns that don’t set them apart. What you really want to do is make IT easier to consume and less complicated, so companies don’t have to spend too much time focusing on it. And at the infrastructure-as-a-service level that’s what’s happened: running systems on instantly accessible servers, storage and networking so you don’t waste a lot of time investing in racking and stacking and configuring them.
As to what’s different for startups versus enterprises, I think startups are much more focused on how they can get resources up and running quickly and get to market fast. At the enterprise level, that’s also happening, and it sometimes takes the form of so-called “Shadow IT.” But there’s another set of requirements in an enterprise where IT departments have to grapple with projects that have been started by developers or business managers, and they need to provide governance over what’s being developed and deployed in order to achieve compliance and security and cost control. That’s when they start looking for some way to manage the ‘runaway train’ of cloud.
GC: Is there a limit to cloud applicability? Is there ever a point where an enterprise is going to be 100 percent cloud? Or is it easier for them always to own?
MC: For today, cloud isn’t the answer for everything. Two examples that come to mind are: 1) laws and policies that don’t accommodate cloud approaches, and 2) software architectures built for traditional IT deployments. Not everything in the world evolves at the same pace. There are aspects of regulatory compliance that haven’t kept up with cloud, and are still catching up. PCI-compliance is just now adapting to the notion of cloud deployments. There are still projects running COBOL code on mainframes that aren’t likely to transition over to cloud anytime soon. All of these will get replaced at some point in the future, but it will take time.
The better perspective on cloud applicability is to look for the low-hanging fruit. What are the projects that are no-brainers to move over to cloud initially and then expand from there? The focus we have here at RightScale is to help customers identify projects and systems that make sense to move to cloud today where there’s a big bang for the buck, and the switching costs aren’t high. New projects are an ideal fit from that point of view.
Check back Thursday for the conclusion of this interview!
By Jake Gardner