Monday, January 10, 2011

Does Gartner understand cloud computing?

I like the Gartner Magic Quadrant. It provides for a quick overview of a specific domain, lists the key players and assigns them a high-level effectiveness rank. It's a great tool, and also a very dangerous one. For those who understand a domain it provides reenforcement and a third-party perspective. And for those who don't, it affords a quick way to select a vendor or choose a technology -- the modern equivalent of "no-one gets fired for selecting IBM".

Having placed many IaaS bets as a cloud management vendor, the Magic Quadrant for IaaS was something I was looking forward to review. But even at first glance it seemed a bit strange, with the title "Cloud Infrastructure as a Service and Web Hosting". Hmmmm... So this is a review of IaaS, the very epitome of cloud services, and also of traditional web hosting services?

Reading through the report you will find similar idiosyncrasies and an obvious bias towards delivering 'enterprise' cloud services:
  • Security is called out multiple times, with no definition of what a secure cloud platform entails. Oddly enough, the paper doesn't call out Amazon as a PCI-DSS compliant vendor, which is one of the clearer methods to show that one vendor is more secure than another.
  • Availability is a key requirement, preferably at 100% uptime. But while availability is important, cloud IaaS is not about giving you an always on website. Rather, cloud is about giving you the tools to make robust web applications that can easily manage with a failed node. If you need a 100% uptime web hosting solution then you should probably look for comparisons between the more traditional web hosts.
  • Professional services are oddly enough a key component of the IaaS cloud. You see, apparently it's not cloud IaaS if it doesn't come with a sales and consulting team to help you deploy it.
Security, availability and professional services --- Gartner is clearly responding to dot-points mentioned to them by the large corporates that consume their material. And I daresay that these companies may not be needing cloud IaaS, but just want to be part of the hype.

You can see the oddity of it all in the actual Magic Quadrant (below). Amazon, considered to be miles ahead in features, mind-share and capabilities is listed as sub-par in "ability to execute". Whilst your friendly corporate vendors, like AT&T and Verizon Business, show up as leaders of the group.

Many large companies practice what I like to call Architecture by Gartner, being quickly magnetized to whichever solutions appear in the Leader's Quadrant. So while I don't know who will be the winner of cloud IaaS for 2011, I can tell you that Savvis, AT&T, Rackspace, Verizon and Terremark will at least be having a very pleasant start to the year.


  1. I'll point out that the Magic Quadrant text says: "Amazon is a thought leader; it is extraordinarily innovative, exceptionally agile and very responsive to the market. It has the richest cloud IaaS product portfolio, and is constantly expanding its service offerings and reducing its prices."

    Indeed, we go on to say many nice things about Amazon as a pure cloud provider.

    On the other hand, despite Rackspace coming out in the overall scoring as a leader, we say: "Rackspace's cloud IaaS is not yet enterprise-class; it significantly lags behind its two most significant competitors (Amazon has a superior feature set, and GoGrid has better features and a full suite of managed services)."

    So you really do have to delve into the text, not just look at the graphic. (Yes, I know that some people will just look at the boxes. They make us sigh with frustration, every single time.)

    I'm planning to write a blog post about why we have cloud IaaS as part of hosting (although part of the obvious answer to that is "every single vendor on the MQ, other than Amazon, is a Web hoster"). But for the moment, I would say you are entirely right that the MQ reflects what Gartner's IT buyer clients care about -- security, availability, managed and professional services, and so forth.

    Pure cloud is conceptually neat, but it's not what our clients are buying, and the MQ is specifically intended to be useful to our clients. It's not an abstract theory of what we think the market should be.

  2. And the full reply:

    Thanks for the commentary; hope the clarification helped.