At this morning’s Bunker Session, the central question of the relationship between cloud computing and open-source software was answered early and often. The one thing everybody agreed upon is that most clouds of any appreciable scale are built on open-source software and, in fact, might not even exist without it. As to whether there’s any money to be made with open source, however, there was enough contention to go around.
The rub, as posited by Citrix’s Simon Crosby, is that everybody making money with open source actually has a proprietary angle. Open source is a great tool for advancing products, branding a company and expanding its reach, but vendors make their money with proprietary solutions. This holds true for companies ranging from Citrix with XenServer to Amazon with EC2. The numbers actually back up this proposition: Jim Zemlin, executive director of the Linux Foundation, pointed out that the leading open-source investors aren’t VCs, but large IT companies like IBM, Intel and Google. Investment in open-source projects helps these companies crowdsource R&D — which saves time and money — before rolling out a commercial offering based on the results. This isn’t an indictment of vendors’ use of open source, by the way, it’s just reality –- not to mention smart business.
However, as Om pointed out from the crowd, there is a big difference between helping vendors make money and actually making money yourself. Save for Red Hat, most truly open-source companies don’t last too long before they’re snatched up by proprietary vendors that want to leverage the associated momentum and product capabilities in their own businesses. But the definition of success isn’t universal. SugarCRM CEO Larry Augustin countered this argument with the position that open-source companies like JBoss and SpringSource, which did great things and built huge communities, are no less successful because they exited via acquisition rather than IPO.
Despite debate over what constitutes a viable open-source business model, the area from where we can expect to see one emerge is cloud computing. As I discussed recently in a GigaOM Pro column (sub. req’d), open-source products are building momentum in private-cloud settings especially, and the reason might be that they help users achieve the same efficiencies as large IT companies that invest in open-source projects.
Another possible avenue for open source success in the cloud is interoperability. When Yahoo’s Tom Hughes Croucher asked the Bunker crowd what the most likely solution to lack of interoperability among public clouds is, open source won by a landslide over vendor-developed standards. Opscodes’s Jesse Robbins buttressed this opinion by pointing out that many concerns over cloud interoperability and application portability can be addressed in the planning stages and by utilizing automation capabilities from companies like RightScale. Another possibility is to use an open-source interface like libcloud, which simplifies movement between clouds.
Where the money comes into play is when CIOs demand these types of capabilities before moving to cloud delivery models. As Accenture’s Joe Tobolski noted, they want to leverage the cloud, but they want to know there’s a workable Plan B and Plan C in case the primary cloud provider goes down. Cloud platforms don’t necessarily need to be open source if they’re open enough to work with third-party managements solutions. If an open-source company can build a cross-cloud automation product that lets businesses tweak it to suit their specific needs, that company might find itself in a position like Red Hat did when users were searching for a viable alternative to Windows.
Check out response to the event on Twitter via #fosscloud.
Full event video and post-game analysis are available to GigaOM Pro subscribers. For a limited time, subscribe using the coupon code BUNKER0310 to get 20 percent off of our already discounted, charter-year price of $79.