![]() |
![]() |
||||||||||||
![]() |
|||||||||||||
[BACK] | |||||||||||||
Does the Positioning of Lutris’s
|
CustomWare is Lutris' newest
Gold Bar Program Partner! Lutris welcomes Salil and Owen's commentary,
reflecting their own opinion.
|
Just when we thought there were too many application servers on the market, Lutris released their Enhydra Application Server (EAS) – previously an open source project, now a commercial application server.
Lutris
feels that their application server is perfectly positioned for the “bottom
two-thirds of the pyramid.” (See Figure 1). Their claim is that the top
third of the pyramid, the Fortune 2000 companies, may legitimately eliminate
EAS as a solution, but that it is perfect for the bottom two thirds.
Is this positioning valid? Should we, the J2EE user community, buy this sales pitch? That’s the question this article tries to answer.
Our verdict: yes, probably, for the reasons below.
The Global 2000 are involved in the eternal battle to obtain and exploit the latest technologies. In their attempts to modernize and standardize, these giants of spending will invest heavily, and sometimes carelessly, in training, consulting, hardware, and of course, software licensing. During our engagements, we have observed that the decisions as to which application server vendor is chosen is more often influenced by politics, fear, and “what everyone else is doing” rather than accurate cost-benefit analysis. (This is not always a bad thing; sometimes it’s actually appropriate).
Smaller enterprises, on the other hand, are often forced to overcome their fears and seek alternatives to the expensive and established vendors. This is also not a bad thing.
A recent Gartner Group study showed that companies overspent by $1 Billion on application servers in the last couple of years [1]. Comments such as the one by Tracy Milburn on a chat board about what happened at his company, are common: “Sad, but true… They were sold a bill of goods… and spent $1M when they could have spent around $100000 (counting development time).” [2]
Vendor / product |
Estimated Cost / CPU |
iPlanet iAS |
$20,000.00 |
BEA WebLogic |
$17,000.00 |
IBM Websphere |
$12,000.00 |
Borland BES |
$12,000.00 |
Lutris EAS |
$4,495.00 |
With some application servers costing as much as $20,000.00 /CPU, a smaller enterprise is forced to look at alternatives to the dominant giants. The Lutris offering of the Enhydra Application Server is just such an alternative and with its unique alternative to the traditional all or nothing J2EE strategy, is a good fit for those enterprises needing to make the most of their time and money.
Companies such as NEC (BEA’s largest reseller in Japan) also resell Lutris EAS. Why? Because the higher-end products are locked into business models that require large margin pricing, which often makes them inappropriate for small- to medium-sized businesses, and ISVs.
Another aspect that ISVs find valuable is that Lutris EAS offers a service-based architecture; enabling rapid integration of existing code and a gentle migration to the complete J2EE stack of services. New applications can be written to the common standards and deployed within a robust and highly performing J2EE environment at the same time that older non-J2EE applications can be deployed to the same VM! This is unique among application server vendors.
What about cheap or free/open-source application servers? EAS’s value proposition, which includes XMLC, it’s pluggable services architecture, its investors and enterprise customers, and it’s training and consulting partners, make EAS worth its price.
Reason #3: The existence of standards may curb “the network effect”.
In his keynote at JavaOne 2001 (San Francisco), Bill Coleman, the then-CEO of BEA spoke of the “network effect.” It refers to the momentum that a product gathers after gaining a large percentage of market share… people start buying more of the product just because it has large market share, and is perceived as the de facto standards.
However,
the existence of standards such as J2EE, the ongoing tightening of the
specifications, and availability of compatibility test suites, are counterbalances
to this effect.
These counterbalances force the established vendors to prove their price-performance ratios every day, and open the market windows for upstarts like Lutris that provide attractive value propositions.
[1] Gartner Group, Aug 21 2001, PR NEWSWIRE release, http://www.gartner.com/
[2] http://www.theserverside.com/home/thread.jsp?thread_id=8603