When you start to think about Enterprise Management Platform (EMP), few questions pop into the head:
- Do we really need one?
- What is wrong with “legacy” platforms?
- Why no one came up with “new thing” yet?
- Does open source really has an answer?
Do we really need one?
I guess the answer to this question depends on what is it we are talking about when we are saying “Enterprise Management Platform”. If we are talking about piece of software which can “discover” all computers on your network, show them on nice and colorful console and then ping them once in a while to make sure that they still plug-ed in – than the answer is most definitely NO. On the other hand, if we are talking about piece of software which can easily integrate “best of the breed” discovery tool with “best of the breed” reporting tool and than provide you with ability to define what exactly you mean when you are saying that Managed Object XYZ is “in groovy state”, then the answer could be quite different.Basically Enterprise IT Universe consists of so many different things and needs to be managed from so many different “angles”, that it is impossible to put all needed knowledge into single product. So Enterprise Management Platform should not even try to do any “management” but instead should transform into “Enterprise Management Tools Collaboration Platform”. What is wrong with “legacy” platforms? But wait, “That Nice and Great” platform has endless list of options. Surely they can cooperate with any “best of breed” tool you care to throw into it. And “Highly Popular Old Working” platform can do web services. Should not it be enough? What about “Best Mechanical Calculator” platform? It was built around “script everything” concept. If you want to deal with “groovy state” or XYZ object, all you need is to write another script.
Unfortunately, it is not that simple. None of the “legacy” platforms were built with much emphasis on “Ease of Tool Collaboration”. There were no reasons for it back then. Companies which developed these platforms truly believed that they can “boil the ocean” and “cure common cold” of Enterprise Management all by themselves (maybe with few acquisitions). Best of them, did have some sort of APIs to let 3rd parties to do some relatively minor additions which were too “small of a fish” to go after. None of them were too eager to let customers define what he/she wants to manage and how. They had very good reason not to make this kind of customization too simple. All of these companies have huge Customers Support teams and they make as much money (and sometimes even more) by “adapting” their software to customer real needs as they make by selling it. If you add to all this the fact that most of the code in these “legacy” platforms was written 5 or even 10 years ago, it will become increasingly clear that “legacy” platforms is a wrong place to look for “The Answer”. Why no one came up with “new thing” yet? The answer is very simple – money. Big companies which can afford R&D expenses associated with development of a new shiny platform, do not have any incentive to do it. They are mildly interested in making their own platforms a little bit more robust/reliable/scalable, but they have all that “compatibility skeletons in the closet” to deal with so they can not do anything drastic. At the same time, they have no interest at all in making platform easily adaptable to customer needs (what all these folks from Customer Support will do?). Small innovative companies just can not cope with this kind of problems. Let’s face it, there is no money in the plumbing. No one in their right mind will spend millions of dollars to develop something they can not sell. Does open source really has an answer? I guess you will never know for sure until you try. But Open Source have some unique features which can make this kind of project a possibility. We can start with most obvious things, like bunch of highly skilled developers which could join such project for all the different reasons people do join Open Source Projects. Then we can talk about almost endless plethora of existing Open Source Projects which could provide building blocks for such project. It could be anything from something huge like JXTA to something small like Apache Commons. Open Source at the moment reached the “critical mass” when you can build almost anything just by combining existing pieces. The last thing in favor of Open Source is the fact that such project will be vendor neutral. Hopefully other companies will not consider it as a threat. Maybe they even will support it (it did become cool lately to support Open Source Projects).