Today, Dortch revisits the topic in an interview with SearchEnterpriseLinux.com. The cost benefits still exist, but now there are choices to consider other than price. Dortch weighs the pros and cons of IBM WebSphere and JBoss.com middleware and advises IT managers on how to find the best fit for their IT environments.
SearchEnterpriseLinux.com: Your research says that intense competition exists between JBoss and IBM's WebSphere Application Server Community Edition. What do IT managers need to know about the difference between the two?
Michael Dortch: Those are the fastest-growing open source application servers out there. Ever since -- and especially since -- Red Hat bought JBoss, they've been able to position it to be much more competitive with WebSphere Community Edition. Red Hat customers should note that the company has expanded its position as a platform provider, not just an operating system provider. And they're not just expanding their business with an application server but the entire platform stack as well. There's more of a WebSphere angle to what they're doing today. When talking to IBM, they are very hot to go after JBoss.
We've established some of the major players. When vetting their options what else should IT managers know about?
When you talk to IBM about WebSphere Community Edition, they love to point to the fact that it is based on Apache Geronimo [Editor's note: In October 2005, IBM announced Community Edition, a free edition of its WebSphere application server that is based on Geronimo.] However, one thing end users have trouble with regarding Geronimo is that it is a blessing and curse in that it comes with all so many options. It's a modular design and can be configured to do anything but make coffee and doughnuts. The downside of that is users have to decide exactly what it is they want it to do. There can be some heavy lifting and assembly. However, when IBM based the community edition on Geronimo, they took some of that out of users' hands, but not in a restrictive way.
On the other hand, what effect has the Red Hat acquisition of JBoss had on IT managers' day-to-day application server operations?
It was a good acquisition from a technology standpoint, but my concern is that Red Hat Linux and JBoss aren't as tightly integrated as they would have liked them to be by now. I get the impression that JBoss is being run more like a business within a business than as an integrated part of the Red Hat family. When companies have to spend time polishing the rough edges of a merger and make things integrated, every hour they do that is one they don't have to provide to customers. As impressed as I am with them, I still get the impression from people and business leaders I speak with that this may be a problem.
How do you advise end users who are weighing a choice between IBM WebSphere Community Edition and something like JBoss?
If you are the IT guy who likes to innovate but who is focused primarily on getting your work done and making the business run, then go with IBM, and they are probably going to serve you very well.
If you are looking to get away from IBM or some other dominant, traditional vendor -- maybe you're feeling what they call "vendor lock-in" -- then you should seriously consider JBoss and Red Hat. But you must understand that you are going to pay something for this enterprise-class support and services regardless of whether or not you go with IBM or Red Hat or their partners. The fact that JBoss or Geronimo is open source and freely available does not mean it is going to be free to run your business on it.
If you are an IT manager at a tech-savvy company and have significant resources in-house or you are already a significant player in the open source community that supports Geronimo and Apache, etc., and basically your business won't die if all of a sudden something has to be brought down for changes, then by all means experiment directly with open source applications.
The good news is now IT managers have a spectrum of choices. It's not open source versus closed anymore. Users can now buy applications that are effectively open inside but have wrapped around them familiar, proven support and services.
You've written that even though the application servers on Linux still save you money over proprietary alternatives, there's more to it than price. What's a concern beyond price?
Support. A company like Red Hat has more to support today. This is not quite a linear platform; it's not just Linux plus JBoss. They want to sell this as an integrated platform, and in a way this solves my business problem if I were an IT manager. To an IT manager, the acquisition makes sense, and it makes sense that Red Hat is now assuming more of the heavy lifting with application servers than they were before. Many IT guys were probably used to this already, since many Red Hat users were JBoss users to begin with. Now they have a choice, where everything is being delivered through one funnel. Enterprise IT guys love the "one throat" mentality, and now Red Hat is that throat.
When you listen to users talk today, the dialogue has changed from TCO to integration, interoperability and enterprise-class support and services. They need these things to sleep at night and to make credible arguments to their skeptical higher-ups and colleagues that these open source/Linux application servers are ready for business. Frankly, Novell and Red Hat are two choices today that are pretty experienced at delivering services and support into IT environments.
Open source application servers have a specialized community of developers built around them. How do you advise clients about participation and allocating resources?
Enterprises today are increasingly borrowing things that were first proven in the open source world and putting them into mission-critical environments. The concept of open source has matured on three fronts: community, standards and technology. If we look at how enterprises are evolving, they are borrowing a lot of the work of the open source communities and adapting it to their own needs.
My best advice about how to prepare for this is an anecdote about the Christian Science Monitor. This organization will give access to open source developers and allow them to meet in the CSM campus [in Boston]. What [CSM] gets in exchange is a chance to interact with this community and direct access to the projects they are developing. This is not just a user group but a developer group, and the CSM can plug their work into wherever it makes the most sense.
In your research, you argue against a total-cost-of-ownership strategy and instead promote a "time to success" model. How have things changed?
Many IT guys today are worrying, "Will proprietary technology A interoperate with type B, and will vendor X's standards work with vendor Y's?" What a lot of our clients are finding today, however, is that, with Linux and open source application servers, the wisdom of the crowd has solved those standards-based problems. And because the management tools are freely available and malleable, they find that the big sell was "cheap and free" but is now "What will complete my processes faster?" Open source used to be inexpensive and free, but now IT is finding value [in it] because it's more responsive.
Email Jack Loftus with your comments and suggestions.