Monthly Archives: September 2013
Massimo Re Ferre’ recently posted some thoughts as a follow-up to his talk at VMworld, about vCHS vs. AWS. That led to a Twitter exchange that made me think that I should highlight a viewpoint of mine:
I do not believe in a “world of two clouds”, where there are cloud IaaS offerings that are targeted at enterprise workloads, and there are cloud IaaS offerings that are targeted at cloud-native workloads — broadly, different clouds for applications designed with the assumption of infrastructure resilience, versus applications designed with the assumption that resilience must reside at the application layer.
Instead, I believe that the market leaders will offer a range of infrastructure resources. Some of those infrastructure resources will be more resilient, and will be more expensive. And customers will pay for the level of performance they receive. There’s no need to build two clouds; in fact, customers actively do not want two different clouds, since nobody really wants to shift between different clouds as you go through an application’s lifecycle, or for different tiers of an app, some of which might need greater infrastructure resilience and guaranteed performance.
I do not believe that application design patterns change to be fully cloud-native over time. First, enterprises have hundreds if not thousands of existing legacy applications that they will need to host. Second, enterprises continue to write non-cloud-native apps, because the typical app is small — it’s some kind of business process app (I call these “paperwork” apps, usually online forms with some workflow and reporting), and it runs on a tiny VM, has few users. It’s neither cost-effective to spend the developer time to make these apps resilient, nor cost-effective to distribute them. Putting them on decently resilient infrastructure is less expensive. Some of these apps should more logically be written on a business process management suite or PaaS (BPMS or bpmPaaS), or on a more general PaaS; that underlying BMPS/PaaS should hopefully functionally provide resilience, but that won’t deal with the existing legacy apps, so there’ll continue to be a need for resilient infrastructure.
When people talk about infrastructure resilience, they’re generally referring to compute resilience in particular — essentially, trying to protect the application from the impact of potential server hardware failure. VMware pioneered two technologies in this space — they call them “HA” (fast detection of physical host failure and automatic restart of the VMs that were running on that host, on some other host) and “vMotion” (live migration of VMs from one physical host to another). However, all the other major hypervisors have now incorporated these features. There’s absolutely no reason why a cloud IaaS provider like AWS, which doesn’t currently support these capabilities, can’t add them, and charge a premium for these VMs.
When people talk about performance consistency, they’re generally referring to storage and network performance. (Most cloud IaaS providers do not oversubscribe either CPU or RAM resources.) Predictable storage performance is a very difficult engineering problem. Companies like SolidFire are offering all-SSD storage to help accomplish this (since it reduces the variability of seek times), and we’re seeing gradual uptake of this technology into cloud IaaS providers. AWS has done “provisioned iops” (PIOPS), allowing customers to buy into a more predictable range of storage performance. There’s no reason why providers wouldn’t offer this kind of predictability for both storage and network — especially when they can charge extra for it.
Now, there are tons of service providers out there building to that world of two clouds — often rooted in the belief that IT operations will want one thing, and developers another, and they should build something totally different for both. This is almost certainly a losing strategy. Winning providers will satisfy both needs within a single cloud, offering architectural flexibility that allows developers to decide whether or not they want to build for application resiliency or infrastructure resiliency.
For more on this: I’ve covered this in detail in my research note, Market Trends: Public and Private Cloud Infrastructure Converge into On-Demand Infrastructure Fabrics (Gartner clients only).
Bernard: “What skill or insight has allowed AWS to create an offering so superior to others in the market?”
AWS takes a comprehensive view of “what does the customer need”, looks at what customers (whether current customers or future target customers) are struggling with, and tries to address those things. AWS not only takes customer feedback seriously, but it also iterates at shocking speed. And it has been willing to invest massively in engineering. AWS’s engineering organization and the structure of the services themselves allows multiple, parallel teams to work on different aspects of AWS with minimal dependencies on the other teams. AWS had a head start, and with every passing year their engineering lead has grown larger. (Even though they have a significant burden of technical debt from having been first, they’ve also solved problems that competitors haven’t had to yet, due to their sheer scale.)
Many competitors haven’t had the willingness to invest the resources to compete, especially if they think of this business as one that’s primarily about getting a VM fast and that’s all. They’ve failed to understand that this is a software business, where feature velocity matters. You can sometimes manage to put together brilliant, hyper-productive small teams, but this is usually going to get you something that’s wonderful in the scope of what they’ve been able to build, but simply missing the additional capabilities that better-resourced competitors can manage (especially if a competitor can muster both resources and hyper-productivity). There are some awesome smaller companies in this space, though.
Bernard: “Plainly stated, why hasn’t a credible competitor emerged to challenge AWS?”
I think there’s a critical shift happening in the market right now. Three very dangerous competitors are just now entering the market — Microsoft, Google, and VMware. I think the real war for market share is just beginning.
For instance, consider the following, off the cuff, thoughts on those vendors. These are by no means anything more than quick thoughts and not a complete or balanced analysis. I have a forthcoming research note called “Rise of the Cloud IaaS Mega-Vendors” that focuses on this shift in the competitive landscape, and which will profile these four vendors in particular, so stay tuned for more. So, that said:
Microsoft has brand, deep customer relationships, deep technology entrenchment, and a useful story about how all of those pieces are going to fit together, along with a huge army of engineers, and a ton of money and the willingness to spend wherever it gains them a competitive advantage; its weakness is Microsoft’s broader issues as well as the Microsoft-centricity of its story (which is also its strength, of course). Microsoft is likely to expand the market, attracting new customers and use cases to IaaS — including blended PaaS models.
Google has brand, an outstanding engineering team, and unrivaled expertise at operating at scale; its weakness is Google’s usual challenges with traditional businesses (whatever you can say about AWS’s historical struggle with the enterprise, you can say about Google many times over, and it will probably take them at least as long as AWS did to work through that). Google’s share gain will mostly come at the expense of AWS’s base of HPC customers and young start-ups, but it will worm its way into the enterprise via interactive agencies that use its cloud platform; it should have a strong blended PaaS model.
VMware has brand, a strong relationship with IT operations folks, technology it can build on, and a hybrid cloud story to tell; whether or not its enterprise-class technology can scale to global-class clouds remains to be seen, though, along with whether or not it can get its traditional customer base to drive sufficient volume of cloud IaaS. It might expand the market, but it’s likely that much of its share gain will come at the expense of VMware-based “enterprise-class” service providers.
Obviously, it will take these providers some time to build share, and there are other market players who will be involved, including the other providers that are in the market today (and for all of you wondering “what about OpenStack”, I would classify that under the fates of the individual providers who use it). However, if I were to place my bets, it would be on those four at the top of market share, five years from now. They know that this is a software business. They know that innovative capabilities are vitally necessary. And they know that this has turned into a market fixated on developer productivity and business benefits. At least for now, that view is dominating the actual spending in this market.
You can certainly argue that another market outcome should have happened, that users should have chosen differently, or even that users are making poor decisions now that they’ll regret later. That’s an interesting intellectual debate, but at this point, Sisyphus’s rock is rolling rapidly downhill, so anyone who wants to push it back up is going to have an awfully difficult time not getting crushed.
Bernard Golden recently wrote a CIO.com blog post in response to my announcement of Gartner’s 2013 Magic Quadrant for Cloud IaaS. He raised a number of good questions that I thought it would be useful to address. This is part 1 of my response. (See part 2 for more.)
(Broadly, as a matter of Gartner policy, analysts do not debate Magic Quadrant results in public, and so I will note here that I’m talking about the market, and not the MQ itself.)
Bernard: “Why is there such a distance between AWS’s offering and everyone else’s?”
In the Magic Quadrant, we rate not only the offering itself in its current state, but also a whole host of other criteria — the roadmap, the vendor’s track record, marketing, sales, etc. (You can go check out the MQ document itself for those details.) You should read the AWS dot positioning as not just indicating a good offering, but also that AWS has generally built itself into a market juggernaut. (Of course, AWS is still far from perfect, and depending on your needs, other providers might be a better fit.)
But Bernard’s question can be rephrased as, “Why does AWS have so much greater market share than everyone else?”
Two years ago, I wrote two blog posts that are particularly relevant here:
- Common Service Provider Myths About Cloud Infrastructure
- In Cloud IaaS, Developers are the Face of Business Buyers
These posts were followed up wih two research notes (links are Gartner clients only):
- New Entrants to the Cloud IaaS Market Face Tough Competitive Challenges
- How Buyers Purchase Cloud IaaS
I have been beating the “please don’t have contempt for developers” drum for a while now. (I phrase it as “contempt” because it was often very clear that developers were seen as lesser, not real buyers doing real things — merely ignoring developers would have been one thing, but contempt is another.) But it’s taken until this past year before most of the “enterprise class” vendors acknowledged the legitimacy of the power that developers now hold.
Many service providers held tight to the view espoused by their traditional IT operations clientele: AWS was too dangerous, it didn’t have sufficient infrastructure availability, it didn’t perform sufficiently well or with sufficient consistency, it didn’t have enough security, it didn’t have enough manageability, it didn’t have enough governance, it wasn’t based on VMware — and it didn’t look very much like an enterprise’s data center architecture. The viewpoint was that IT operations would continue to control purchases, implementations would be relatively small-scale and would be built on traditional enterprise technologies, and that AWS would never get to the point that they’d satisfy traditional IT operations folks.
What they didn’t count on was the fact that developers, and the business management that they ultimately serve, were going to forge on ahead without them. Or that AWS would steadily improve its service and the way it did business, in order to meet the needs of the traditional enterprise. (My colleagues in GTP — the Gartner division that was Burton Group — do a yearly evaluation of AWS’s suitability for the enterprise, and each year, AWS gets steadily, materially better. Clients: see the latest.)
Today, AWS’s sheer market share speaks for itself. And it is definitely not just single developers with a VM or two, start-ups, or non-mission-critical stuff. Through the incredible amount of inquiry we take at Gartner, we know how cloud IaaS buyers think, source, succeed, and sometimes suffer. And every day at Gartner, we talk to multiple AWS customers (or prospects considering their options, though many have already bought something on the click-through agreement). Most are traditional enterprises of the G2000 variety (including some of the largest companies in the world), but over the last year, AWS has finally cracked the mid-market by working with systems integrator partners. The projected spend levels are clearly increasing dramatically, the use cases are extremely broad, the workloads increasingly have sensitive data and regulatory compliance concerns, and customers are increasingly thinking of AWS as a strategic vendor.
(Now, as my colleagues who cover the traditional data center like to point out, the spend levels are still trivial compared to what these customers are spending on the rest of their data center IT, but I think what’s critical here is the shift in thinking about where they’ll put their money in the future, and their desire to pick a strategic vendor despite how relatively early-stage the market is.)
But put another way — it is not just that AWS advanced its offering, but it convinced the market that this is what they wanted to buy (or at least that it was a better option than the other offerings), despite the sometimes strange offering constructs. They essentially created demand in a new type of buyer — and they effectively defined the category. And because they’re almost always first to market with a feature — or the first to make the market broadly aware of that capability — they force nearly all of their competitors into playing catch-up and me-too.
That doesn’t mean that the IT operations buyer isn’t important, or that there aren’t an array of needs that AWS does not address well. But the vast majority of the dollars spent on cloud IaaS are much more heavily influenced by developer desires than by IT operations concerns — and that means that market share currently favors the providers who appeal to development organizations. That’s an ongoing secular trend — business leaders are currently heavily growth-focused, and therefore demanding lots of applications delivered as quickly as possible, and are willing to spend money and take greater risks in order to obtain greater agility.
This also doesn’t mean that the non-developer-centric service providers aren’t important. Most of them have woken up to the new sourcing pattern, and are trying to respond. But many of them are also older, established organizations, and they can only move so quickly. They also have the comfort of their existing revenue streams, which allow them the luxury of not needing to move so quickly. Many have been able to treat cloud IaaS as an extension of their managed services business. But they’re now facing the threat of systems integrators like Cognizant and Capgemini entering this space, combining application development and application management with managed services on a strategic cloud IaaS provider’s platform — at the moment, normally AWS. Nothing is safe from the broader market shift towards cloud computing.
As always, every individual customer’s situation is different from another’s, and the right thing to do (or the safe, mainstream thing to do) evolves through the years. Gartner is appropriately cautionary when it discusses such things with clients. This is a good time to mention that Magic Quadrant placement is NEVER a good reason to include or exclude a vendor from a short list. You need to choose the vendor that’s right for your use case, and that might be a Niche Player, or even a vendor that’s not on the MQ at all — and even though AWS has the highest overall placement, they might be completely unsuited to your use case.