Despite growth in open source, commercial solutions still dominate
Open-source cloud computing project OpenStack turned one this week, and cloud backup provider Backblaze freely shared detailed specifications for a storage device capable of holding 135 terabytes of data. Open-source options for everything from servers and data centers to clouds and application platforms are growing more robust. And yet both new and established commercial providers continue to sell products and services. Despite the ready availability of free solutions, the majority of customers continue to prefer products backed by the expertise and support of commercial organizations.
Launched a year ago, OpenStack has already reached a state of maturity sufficient for GigaOM’s Derrick Harris to suggest that it “looks not only like an open-source alternative to Amazon Web Services and VMware vCloud in the public Infrastructure as a Service space, but also a democratizing force in the private-cloud software space.” Eighty-nine companies are now involved, Rackspace is transitioning its existing cloud-computing offering to the OpenStack Compute code, and competitors such as Dell and Internap are preparing their own OpenStack powered services for launch.
Backblaze’s work is different. It is less about a community effort and more about the work of a single company. Nevertheless, it is a good example of an open-source solution that challenges the economics behind existing proprietary approaches. Backblaze’s figures suggest that Amazon could charge almost $2.5 million to store a petabyte for three years and that Dell hardware would cost just over $500,000 to buy and run. But the Backblaze design would cost less than $100,000 over the same period. The company’s founder and VP of engineering, Tim Nufire, is, however, quick to stress that “the economies of scale only kick in if you really do need to store a full petabyte or more.” By taking an approach that lowers the cost of storage, Backblaze has enabled third parties such as Vanderbilt University to consider projects that previously might never have begun.
OpenStack and Backblaze are just two parts of the continuing wave of open-source innovation; other examples include projects such as Eucalyptus, Nimbula, Cloud Foundry, Hadoop and others. Although the fruits of these projects are freely available for anyone to implement, the reality remains different. Especially in mission-critical business areas, many companies remain uncomfortable about downloading plans or compiling code for themselves. Instead, they turn to third parties to help deploy these freely available tools, and they take out support contracts to ensure that someone is always available should things go wrong. Red Hat demonstrated this model almost 20 years ago, selling support around the freely available Linux operating system. Eucalyptus continues in almost exactly the same manner today, and in the OpenStack community we are already seeing the emergence of commercial support from Rackspace Cloud Builders and others. And while some of these third parties (Eucalyptus Systems, for example) are newcomers established specifically for this purpose, many are the existing hardware vendors and systems integrators from whom enterprise IT departments have bought for decades.
But rather than stand idly by and allow open-source alternatives to cannibalize their revenues, established providers such as Dell, IBM and Citrix are finding ways to engage with and develop on top of open-source activity. For example, Citrix, with OpenStack and Project Olympus, is building on top of open-source code in order to create new commercial offerings that benefit directly from the community’s work. Last week’s acquisition of Cloud.com, however, may have altered the detail of this approach, as a recent weekly update explored.
The cost savings that are available to end users willing to sacrifice support and accountability are equally relevant to providers of commercial software and services today. By building their next generation of products on top of open-source foundations, these companies lower their own costs, gain access to innovation from a community of developers for whom they do not need to pay and — potentially — innovate faster and more affordably. Commercial success and open source do not need to be opposites, and the partnership between the two will increasingly be seen as mutually beneficial.
Hi Wendy. Thanks for this. Part of the advantage of having commercial vendors get on board to offer their own take on the underlying OpenStack code is that we may well find early adopters simply buying an OpenStack solution from their existing vendor for other things?
I’ll personally be interested to see how far each of these commercial offerings seeks to differentiate itself through value added services that don’t modify the code, and how far they differentiate by actively adding code that’s unique to themselves. And *if* Citrix (for example) adds significant tweaks to their own flavour of OpenStack, will the consultants I’ve hired from Cloud Builders (for example) refuse to touch it?
Paul, good story- wish this was public so the discussion could be had with others. What I find interesting about OpenStack is it’s not “opensource” like Redhat- it’s an ecosystem play loosely held together by the hope of vendors that people will adopt this way of buying. It’s not a platform that the development community is building against, it’s a grouping of offerings that commercial vendors are building against. So it will be interesting to see if Dell and Citrix plan to offer this ( with their own value add ), just as you mention above, who will support it ( rackspace or the commercial provider ). I personally think it feels like these offerings will stall adoption as IT staff to figure out what it all means- again. And as you say, determine what opensource is okay within IT and where it isn’t okay. Your question of the week is right on target! Can’t wait to hear the results.