Latest appserver market analysis: BEA 41%, IBM 31%, IPlanet 13%


News: Latest appserver market analysis: BEA 41%, IBM 31%, IPlanet 13%

  1. Independent research by Gartner Dataquest shows that BEA holds 41% of the new direct license revenue in the application server software market in 2000, a figure that is 10 share points higher than its closest competitor.
    BEA now has confirmation of its first place market position by every major industry analyst firm tracking the application server market.

    The Gartner market analysis revealed the following market share results:

    BEA - 41%
    IBM - 31%
    iPlanet - 13%
    Oracle - 4%
    HP Bluestone - 4%

    Full release

    Threaded Messages (21)

  2. '..holds 41% of the new direct license revenue in the application server software market...'

    I assume then that any open source (free) server will forever rank at 0% on this survey? Nice.
  3. ...yeah, Gartner completely blows. I noticed their latest article "Open-Source Application Servers: A Viable Alternative?"
    recommends Lutris, which is no longer OpenSource and then mentions JBoss, the most successful and relevent Open Source app server in a footnote--after Apache which, albeit OpenSource, is not an app server.

    What Gartner does not know about app servers could fill volumes.
  4. Actually, as far as I can tell Lutris never actually released the source, which seems to me to be a pretty large part of the 'Open SOURCE' idea.
  5. Hey , Can any one tell me , how is Borland AppServer.
    According to the analysis seams to have 0% market share.
    I have to use it in my next project . it doesn’t seam to be that
  6. Borland's app server is fine. It seems to work well, particularly for development purposes. It isn't a market leader, and it isn't free, so it is stuck in market nightmare land, at no real fault to itself. Don't dismiss it out of hand, but try to stick to the J2EE standards so if you end up deploying on something else it will minimize the pain.


  7. Sushant, We have been using Borland Appserver for several applications. The user base is about 200 internal users(but distributed accross the country). We have had a lot of success with BAS and JBuilder Enterprise. Prior to this we were using Visbroker and our new applications continue to integrate well with those applications that use Visibroker CORBA. All in all the only problem you may find with Borland is their support or lack thereof both from a sales and technical perspective. Though the technical support seems to have improved this year.

    Feel free to email me with any questions

    Savio Lobo
  8. Borland AppServer[ Go to top ]


    As the prsvious writers said, BAS is fine. It is IMO the best server for development, technically speaking it is also the best server for deployment (IMO again), though the decision may be more politically there.
    Basically I'd say BAS is an outstanding implementation of the J2EE, but it does not offer much "added value" not included in J2EE.
    IMO this is nothing bad, just the other way round, as I like to stick to the standards. Probably everything other vendors are implementing in their servers "proprietary" could be built on top of J2EE (as e.g. SilverStream proves), so nothing to worry about unless you like vendor lock-in *gg*
    BAS is also rated very high by people who used it and if Borland is clever it will gain more marketshare in the future, though it will probably never be a market leader.


  9. How about 2001 market? I heard IBM have gained much more with thier launch of WebSphere 4...

    Wei Boon
  10. How about 2001 market? I heard IBM have gained much more with thier launch of WebSphere 4... <

    There will be a 'blip' as they get a few new purchases that were put off until the WS4 release and perhaps from upgrade revenue etc.

    However, WS4 is not yet commercially viable IMHO. (We develop for WS4 all the time. It is still unusable.)


  11. Ref: Cameron - I was just wandering how you find WebSphere 4 unusable. We are starting to use it on a project and so I would be very interested to hear why you think it is not practicle.


  12. I wonder how much BEA paid Gartner to place their appserver as number one !! how about that for a question?

    I've seen big companies pay big money to have the leading percentage in any statistics.

  13. Hi M&M,

    I doubt any of those reputable research MNCs would jeopardize their names just to earn a few measly bucks. If they did, word would have gotten out since Day One and they would have been shot out of the research industry long ago.

    The vendors can pay these benchmarkers to initiate a research project but their results are seldom swayed. Too much reputation and integrity at stake.

    Giga, Gartner, and Meta have all posted similiar results. I doubt vendors would be willing to fork out millions just to engage in marketing wars.... Hang on... I think IBM just did that but that's another story altogether. ;)

    Anyway, I'm sure they'd rather use the money for R&D to improve their products, which is better for us developers.

    J2EE rules!
  14. Ref: Cameron - I was just wandering how you find WebSphere 4 unusable. <

    When we encounter a problem in Websphere, it takes us days to resolve it. We encounter problems often. There are many annoying little bugs with no work-arounds and there are significant architectural flaws, particularly regarding class loaders. (The system classloader does not load all java and javax packages for example!)

    The J2EE standards have been retrofitted onto Websphere. That means that Websphere does NOT run J2EE applications. Rather, you write Websphere applications that just happen to occasionally implement J2EE interfaces and use J2EE APIs.

    Once you get it up and running, and have your app operating, it does seem to work for the most part. However, unlike other IBM products, that is not a guarantee. Websphere can suffer unexplained catastrophic failure out of the blue requiring complete re-installs. Ours is not the only company to witness this!

    So you might think that I hate Websphere. Not at all. It's not an emotional response, it's just that we can much more efficiently develop for, support, etc. other platforms.

    My personal favorites are Resin (for web container apps), Orion (for "departmental" applications running on a single server) and Weblogic (for high scale apps and some nice technical features).


  15. I just wonder how this is measured... in terms of bought licenses? Usage?
    Simply because I've seen some projects started by untrained people, AppServer bought, project fails.
    Then they call someone who knows the technology, AppServer is often replaced, and so they have an AppServer license never used.
    Just wondering...

    kind regards

  16. Yeah, that's the great thing about surveys and analyst groups -- it's all relative (translated: skewed to fit some purpose). You know that they are not counting any OpenSource solutions in this.

  17. The year 2000 ended 101/2 months ago. The press releease
    from BEA on 10/17/01 shows sheer desperation for good news.
  18. Posted By: Michael Downing on October 23, 2001 in response >to this message.

    >The year 2000 ended 101/2 months ago. The press releease
    >from BEA on 10/17/01 shows sheer desperation for good >news.

    The Gartner Dataquest report is dated October 2001.
    "Source : Gartner Dataquest "The Application Server Market Will Stay Healthy", J. Correia, October 2001."

    However, these figures are similar to a Meta Group survey which was conducted in September 2001 (comissioned by BEA)

    This reports
    North American market shares of application server vendors supporting J2EE are:
    o BEA WebLogic &#8211; 37%
    o IBM WebSphere &#8211; 22%
    o Oracle Application Server &#8211; 11%
    o iPlanet Application Server &#8211; 5%
    o Others &#8211; 12%

    North American market shares for application server vendors supporting EJB are:
    o BEA WebLogic &#8211; 52%
    o IBM WebSphere &#8211; 14%
    o Oracle Application Server &#8211; 6%
    o iPlanet Application Server &#8211; 4%
    o Others &#8211; 24%

    More at
  19. The IBM stance on this is interesting...
  20. excuse me, but how 'appserver' is defined? I am always wandering if MS' DNS (is this the name?) should be included? If yes, I think it should have a fairly big market share.
  21. Posted By: ted slusser on October 24, 2001 in response to >this message.

    >The IBM stance on this is interesting... would be interesting if it was a response to the same survey - but it isn't.

    IBM issued the above release on 25 July 2001. The latest BEA releases are from Sept 26 (Meta) and Oct 17 (Gartner).

    Back in July IBM claimed to be "leading in growth" - the figures above don't seem to bear this out.


  22. HP Bluestone 4%


    who is actually using this?