Discussions

News: JBoss in the Cloud

  1. JBoss in the Cloud (29 messages)

    JBoss is announcing the availability of JBoss Enterprise Application Platform (EAP) on Amazon's Elastic Compute Cloud (EC2) as a supported Beta, available through Amazon Web Services. This means developers now have access to pre-built JBoss software appliances, production support with a 2 business-day response SLA, and software update access via Red Hat Network for a fixed per month fee of $119, and a variable fee starting at $1.21 per instance hour. For more info on Red Hat and Cloud Computing, go here.

    Threaded Messages (29)

  2. Re: who cares[ Go to top ]

    Honestly, I love JBoss, I defend them at every turn, but after Muzilla's article in the Register yesterday putting JBoss 5 at "several quarters" out, I question their ability to maintain leadership over Oracle, IBM, or Sun... this is total noise, meaningless noise, i would like to hear how this does anything to change the dynamics of the app server market, other than the supposed marketing benefits of being the 1st app server in the "cloud"... marcf, r u out there, maybe u could write a whitepaper or do some consulting on how to right the ship, because i am getting a bad feeling coming out of NC...
  3. Re: who cares[ Go to top ]

    This is why Bill Burke is so pissed at me, I think - this was the sort of non-announcement RH threw out on a regular basis, and expected me to publish here on TSS.
  4. Re: bored[ Go to top ]

    Joe, Feisty in your emeritus years.... i will regret this comment: but it was 1 thing when JBoss was running the world, pre-Spring, pre-Glassfish, and we could all understand the headiness of the moment when JBoss was establishing OSS middleware... but, to listen to Bill Burke now is just disheartening, i mean isn't he the guy who does RESTful web services, and talks about everything else that provokes an outrage, i don't do much more, but i also don't represent a publicly-traded company with my comments... at some point, maybe over the next "several quarters" while we wait for JBoss 5, Bill should be reigned in, and someone with details that doesn't border on propaganda (read: Sacha over Burke or Muzilla) will step up and start informing the Java community about something that borders on meaningful explanation... until then, the countdown begins for the next Burke meltdown...
  5. Re: bored[ Go to top ]

    Well, I finally get to have an opinion, is all.
  6. Re: bored[ Go to top ]

    but, to listen to Bill Burke now is just disheartening, i mean isn't he the guy who does RESTful web services, and talks about everything else that provokes an outrage, i don't do much more, but i also don't represent a publicly-traded company with my comments...
    Douglas, not sure if you're meaning on TSS or my blog...If you mean TSS, then yes, there isn't much else to do here other than to provoke outrage and incite the zealots and trolls as the site really isn't very interesting and hasn't been for a long time. Its why I pay much more attention to infoq.com nowadays. Question: RESTful web services provokes outrage? REST is propaganda? Maybe I was just misinterpreting your comments. As far as EC2, this may not be big technical news, but pretty interesting from a business sense. Joe seems to ignore our more technical news posts like the one I did a few weeks ago about JBoss "Black Time", our binary drop in replacement for Tuxedo, but that didn't make the cut for some reason... -- Bill Burke JBoss http://bill.burkecentral.com
  7. Re: bored[ Go to top ]

    As far as EC2, this may not be big technical news, but pretty interesting from a business sense. Joe seems to ignore our more technical news posts like the one I did a few weeks ago about JBoss "Black Time", our binary drop in replacement for Tuxedo, but that didn't make the cut for some reason...
    Because I wasn't working on TSS at the time, perhaps?
  8. Re: clarification[ Go to top ]

    Bill, I tried to dissuade u from these unfortunate comments, but u took my bait, and so here it is: i think the clarification is that JBoss has not been very interesting in a long time, i personally think that management matters, and losing marcf, however necessary from an accounting and Red Hat standpoint, seriously hurt the general direction of your org.... i can understand that u think infoq.com has garnered a place in the Java developer's list of referencable sites for technical news, but TSS is still the standard-bearer, and will be as long as these debates are not overrun by propaganda and rants... u r entitled to your viewpoints, but complaining about TSS does not solve the lack of information in the marketplace about JBoss 5, and so u and the middleware group at Red Hat are missing a chance by not using this forum 2 address the larger community here... it just seems like no one has agreed internally at JBoss what the problem is with this release, again, i know it is hard work, but just more candor on what can be expected in the interim between beta 4 and RTM would go a long way in keeping the masses with you guys... berating or even degrading TSS is exactly the opposite of what u should be doing... douglas dooley douglasdooley.blogspot.com
  9. Douglas...[ Go to top ]

    Douglas, Glad to read so much passion in your posts :) I've just posted a blog entry on that very topic, feel free to comment: http://sacha.labourey.com/2008/06/28/jboss-as-50-status/ Onward, Sacha CTO JBoss, a division of Red Hat
  10. Re: who cares[ Go to top ]

    Can I have the URL of this article?
  11. Re: register article link[ Go to top ]

    here it is: http://www.theregister.co.uk/2008/06/16/jboss_5_boast/ ignore my comments, i was a little hard on muzilla, but if you have some thoughts on "re-factoring", I am all ears... is this another way of saying modularization or OSGi?... i know, i know, amateur question....
  12. Automatic scaling?[ Go to top ]

    Getting JBoss to run on EC2 is easy .. its just a virtualized host. The really cool thing about cloud computing is the ability to scale out on demand. I don't think JBoss has cracked that nut yet. We're working on that in WSO2 for the WSO2 Web Services Application Server (WSAS) (http://wso2.org/projects/wsas/java) and then ESB and our other products. The idea is to have a dynamically growing cluster which grows and shrinks as load changes. This is similar to what Google AppEngine gives for Python, but this'll work for Java services (and applications) as well as Javascript and other JVM based languages that are supported on Axis2/WSAS. Afkham Azeez is the person leading this work and he's blogging about it at http://afkham.org/. We'd welcome any help with this work .. all the work is under Apache License 2.0. Sanjiva. http://sanjiva.weerawarana.org/
  13. Re: who cares[ Go to top ]

    this is total noise, meaningless noise, i would like to hear how this does anything to change the dynamics of the app server market, other than the supposed marketing benefits of being the 1st app server in the "cloud"...
    Doug, clouds, grids, etc. are already a required deployment platform for us - so yes - nothing earth shattering - people have been playing with JBoss on EC2 since it was launched. So we're just playing catchup with some of our AlphaGeek customers. Hopefully, this formalizes the process and makes things easier for people who want to try it. And many people do want try it. I think we'll all learn something from the experience and that will gives us the opportunity to improve things. But at the end of the day it's just another platform we *have* to support. Admittedly one that completely changes the economics of infrastructure deployments. Rich Sharples Apprentice JBossian http://blog.softwhere.org/
  14. Rich, I understand it is on the radar of some of your customers, but playing cat-and-mouse with some of my broader contentions, such as that EC2 will never be a deployment platform for enterprises hardly justifies a response... I guess you guys think I am all wet with the more inflammatory arguments, by not responding to these pertinent issues, or u r devising some long-overdue campaign to educate your audiences what can be expected from JBoss 5... But, seriously, I am done. I am going to try and restrain from posting about JBoss for the time being, perhaps until that "several quarters" time-period elapses, and I can comment on something in the marketplace... Unfortunately, whether you value it or not I have been a strong proponent up-to-this point, and I feel like quoting Lauryn Hill: "u just lost another 1..." its a shame, i have no doubt about the prospects of JBoss as a middleware deployment platform for Red Hat customers, an extremely lucrative market in itself, but I am concerned that myopia is creeping in to the operations, when a simple meeting with relevant individuals could have produced some informative responses on this thread... i guess that is 2 much to ask...or perhaps i am a little senstive and/or feeling self-important today... btw, Rich, if you get a chance, in your downtime, check out my latest blog entry and throw me a comment on what u think, as a former Sun guy, I would think u could tell me as well as anyone what u think about the proposal of: Astro Microsystems, at least in theory... douglas dooley douglasdooley.blogspot.com
  15. Re: not what we were looking 4[ Go to top ]

    Rich,

    I understand it is on the radar of some of your customers, but playing cat-and-mouse with some of my broader contentions, such as that EC2 will never be a deployment platform for enterprises hardly justifies a response...

    I guess you guys think I am all wet with the more inflammatory arguments, by not responding to these pertinent issues, or u r devising some long-overdue campaign to educate your audiences what can be expected from JBoss 5...

    But, seriously, I am done. I am going to try and restrain from posting about JBoss for the time being, perhaps until that "several quarters" time-period elapses, and I can comment on something in the marketplace...

    Unfortunately, whether you value it or not I have been a strong proponent up-to-this point, and I feel like quoting Lauryn Hill: "u just lost another 1..."

    its a shame, i have no doubt about the prospects of JBoss as a middleware deployment platform for Red Hat customers, an extremely lucrative market in itself, but I am concerned that myopia is creeping in to the operations, when a simple meeting with relevant individuals could have produced some informative responses on this thread...

    i guess that is 2 much to ask...or perhaps i am a little senstive and/or feeling self-important today...

    btw, Rich, if you get a chance, in your downtime, check out my latest blog entry and throw me a comment on what u think, as a former Sun guy, I would think u could tell me as well as anyone what u think about the proposal of: Astro Microsystems, at least in theory...

    douglas dooley
    douglasdooley.blogspot.com
    Douglas, your concerns for JBoss 5 are valid, but I think you need to review exactly what we've done in the past 1-2 years. Take a look at JBoss Developer's Studio. We put a huge investment into tooling and now we have some very nice eclipse plugins for seam, drools, jbpm, etc... covering most of our product suite. Tooling has long been a weakness of OSS in general, and we've made a huge investment to change this perception. The JBoss Messaging guys have done an incredible amount of work to revamp and revitalize our messaging solution. Full clusterability along with incredible performance have highlighted this. Don't believe me? Propaganda? Go check it out... Then we have Seam and Gavin on Web Beans pushing things. Then you've got JBoss ESB, not revolutionary yet, but considering we had nothing less than 2 years ago, pretty impressive IMO. JBoss Cache guys have been putting out some innovative and quality offerings and are currently working on MVCC locking policies. You have the JBoss Transaction Services team working on cool things like WS-BA and ease-of-use annotations that make it easier to build compensating transactions. They've also acquired a large codebase so that we can provide a binary drop-in replacement for Tuxedo, a.k.a. The "Black Tie" project. Then you have the Ice Tea project, which is a certified full, OSS distribution of Java 6. Hibernate Search, Shards, Validation, and my REST/JAX-RS implementation. All relatively new stuff. You also have Mobicents which is a JAIN/JSLEE platform allowing you to do things like SMS, SIP and fun stuff like that. Finally, we've productized the relevant features of EE 5 in JBoss 4.2/4.3 Platform focusing on giving customers/users what they want as we carefully build our next-gen app server. No, we haven't been good at getting the message out on all these things. A minor part of it was Joe's "objective journalism", but most of it is our developers not blogging enough and not enough reaching out to more reputable communities like infoq.com, onjava.com, or even IEEE. I know Rich has it on his plate to rally the troops to get us communicating more, but if people like yourself Douglas looked into things a bit more that would help a lot as well. Finally, Douglas, I've never had a problem with you, because well, you say what you think and your opinions are your own, and you don't have any personal agendas. I'm sorry if we've lost you, but I think based on customer and user feedback, we've made a majority of good choices on where to place our resources. -- Bill Burke JBoss, a division of Red Hat http://bill.burkecentral.com
  16. ... with some of my broader contentions, such as that EC2 will never be a deployment platform for enterprises ...
    May I ask: Why is that?
  17. Rich,

    I understand it is on the radar of some of your customers, but playing cat-and-mouse with some of my broader contentions
    Doug, don't take this personally - I'm not playing cat and mouse with you; I don't have time. Some advice though - get out of the peanut gallery; become a customer, user or get involved in our communities. You'll learn and your opinion will have a lot more weight.
    btw, Rich, if you get a chance, in your downtime, check out my latest blog entry and throw me a comment on what u think, as a former Sun guy, I would think u could tell me as well as anyone what u think about the proposal of: Astro Microsystems, at least in theory...

    douglas dooley
    douglasdooley.blogspot.com
    I have lot's of opinions of Sun - I'm just trying to resist looking like a bitter and twisted ex employee (which I'm not); though it's a real challenge sometimes ;) - Rich
  18. Re: unfortunate[ Go to top ]

    I write this from a place of discouragement, i honestly don't know what to say about the responses, perhaps i was a bit harsh concerning the plight of JBoss 5, but the coded as well as the not-so-veiled references make me think that some of u guys took the comments personally, 2 bad 4 u... that happens in enterprise software, all the time, people invest their energy and hopes in a highly competitive market, as well as their future livelihoods, and it takes on part of their personality, all this is o.k.... but not addressing anything is fine, i don't really care what happens with JBoss, Inc. at this point, what is unfortunate is the comments that are intentionally condescending, and i don't have the luxury of being anything but in the "peanut gallery", sorry Rich, your understanding of my p.o.v., that i should edit myself, r the same as Burke going off on TSS, its counter-productive, i will take this issue with me, in my travels, and will not bother u guys anymore, the fact that u think that is better than the alternative is your problem, good luck with the to-be-announced launch...
  19. Re: unfortunate[ Go to top ]

    and i don't have the luxury of being anything but in the "peanut gallery"
    Doug, you're missing my point completely. You absolutely do have the luxury. Red Hat (and JBoss) are built on the principals of Open Source. If you don't like what you see - you have the power to do something about it. If you don't like our JDBC connection pool implementation; tell us exactly why or even propose a better alternative. If you don't like the way the products and technologies are marketed - show us how it's done. Open Source can give you that power - but you have to earn the right first - it's a meritocracy. To earn the right - you have to invest - become a customer, user or get actively involved in our communities. It's no accident that many of JBoss' best and brightest and many of its leaders were once customers. So, what do you want to do - throw peanuts or change the world ?
  20. Re: marketing[ Go to top ]

    As some of u know, i take a p.o.v. that is less balanced by the rigors of programming than some others, and that is because i do believe that in a technical marketplace, there is functionality and there is perception, and there is no clear-cut way to argue over which is more important: is Windows functionality easy? yes; is it successful based on perception from users? yes; is 1 more prominent than the other? unclear....and this could be a running dialog ab/ every good piece of software from Oracle DB, WebLogic, OSS middleware of all sorts, and development tools: perception is key to gaining necessary credibility as the organization builds functionality... I have taken some inflammatory pot-shots at the following: JBoss 5, WebLogic as a stand-along entity, Sun on non-Glassfish-related software, WebSphere., Mule on JBI, and SpringSource on anti-EJB rhetoric...I am in the upper-decks looking down on the playing field, not really in the game, and making observations that i hope will achieve the goal that me and some friends talked about in 1999 with the advent of Enterprise Java: release the stranglehold of business application development from Microsoft's grip. so, I get on JBoss org.'s case for the same reasons that I wish Mule would get on the JBI tip, or that we could find some common ground between Spring and EJB, or that Glassfish could be rolled-into a revenue business for Sun: because this opportunity to shape enterprise development is potentially fleeting, and even with a large lead, there is no denying the potential of .Net... so, Rich, I would say to you that i would be happy to concoct some more ideas on what could be done from a marketing standpoint, but u guys don't pay me enough (=0), and i have already pointed out the fuzziness of the "refactoring" discussion from Muzilla in the above Register article, and my points made to Andrew Rubinger ab/ a "massive, coordinated, and expert dissemination" of information is intended to keep people up-to-speed who don't have level 1 support contracts or who want to download betas and dig in 2 the forums... i mean i'll try and give some advice, i just thought that i did already, and OSS ain't charity, i charge for my more-fleshed out product positioning services... sorry for length, JBoss is going to b o.k., all the best, douglas dooley douglasdooley.blogspot.com
  21. Missing the Point[ Go to top ]

    Wow, are you guys off the mark. *JBoss EAP on EC2* This marks a supported entrance into on-demand computing, removing the hardware, OS, and other miscellaneous BS that comes coupled into scope of developer concerns in a traditional hosted environment. To have this prefab deployment structure that maintains its application-level flexibility is more than newsworthy, it's exciting. I can't wait to see the kind of adoption and competition this provokes. *Burke* I'm not sure when passion became an undesirable quality. Some bias is to be expected when forming a argument, and I don't recall any of his writings to be devoid of qualified backing support. Moving on to a new project makes one's opinions irrelevant? *JBoss5* Completely unrelated to the topic of the post. For years the community has been bashing the monolith of the Application Server. JBoss, in turn, has not only listened but done one better by taking steps to modularize and componetize nearly every aspect of its flagship product. This means that throughout the extensive Beta release process, many of the building blocks can be (and are intended to be) used standalone. MicroContainer is doing incredible things. JBoss Messaging. EJB3 Release Candidates. JBoss AOP. Please, use these, try them out. We've got forums and IRC @ Freenode #jboss for early adopters. S, ALR Disclaimer: JBoss Employee; the opinions are mine.
  22. Re: andrew[ Go to top ]

    I am literally holding my hands, trying not to type 'propaganda', so i'll refrain from stating the obvious, but lets talk about a couple points: a. EC2 is a total unknown, maybe some developers will rely on it for independent deployment, but will a single enterprise utilize for in-production deployment? b. It's not exciting, its a potential deployment alternative to Google, thats all...it will not pick up JPMorgan Chase, Caterpillar, or WalMart as anything more than testing environment customers, not their core systems... c. JBoss 5 will be great, no one is really questioning that, and I am sure u r getting hard-core JBoss developers to do the modularization exercise by testing the pieces before the whole, but the impending release of JEE 6 and the day-by-day execution of SpringSource are quickly making it irrelevant in the marketplace for mass adoption - - i hate to even write those words, but that is the reality outside of Raleigh... the only thing that should be going on right now in your organization is massive, coordinated, and expert dissemination of details of what to expect, what the ramp-up will be, what partners are signed on and testing betas, what dates to expect Seam 2.x, Exadel, and SOA integration, I mean just general marketing stuff... Rich Sharples, I apologize, but i think it is time for someone to make a clear statement in the marketplace, beyond what Muzilla said in the Register earlier this week, something that appeases the Java developer community, and says something about the carrots that will lure developers to the 5 platform... i apologize, i apologize, i apologize, just thought someone should b a check 4 some reality, maybe i am alone, but i would think there is some of this shared sentiment out there...
  23. Re: andrew[ Go to top ]

    i apologize, i apologize, i apologize, just thought someone should b a check 4 some reality, maybe i am alone, but i would think there is some of this shared sentiment out there...
    No need to apologize. The problem is that a lot of snake-oil salesmen on all sides are DESPERATELY invested in making sure their snake-oil sells - and they take any doubts as attacks. I'm all into grids - obviously, since I work for a grid company now - and i think it's the future, really. But.... part of the reason I work for Gigaspaces now is because I think they have the right approach to the grid: a shallow curve, one that doesn't necessarily appeal to the guys who want the shiniest, newest thing possible, and who ignore anything that's from yesterday in favor of the newest bauble... but that shallow curve is how grids are going to be made more popular, because the rank and file can actually understand it and apply it. I'm all for innovation - it just needs to be marketed and evangelized for what it is, rather than as the hammer that fits every possible nail. So ... Douglas, good on you, mate - keep up the doubts.
  24. ec2[ Go to top ]

    The thing, almost anything will run on ec2 so long as it doesn't use udp or multicast. I'm pretty saure WebSphere will run there as would any other IBM product unless a prereq check stopped the install because of the white box linux being used. I think the interesting thing about middleware on ec2 is not the fact that it runs there but it's more likely whether things like rightscale etc can be used to scale it and manage it.
  25. snake-oil[ Go to top ]

    The problem is that a lot of snake-oil salesmen on all sides are DESPERATELY invested in making sure their snake-oil sells - and they take any doubts as attacks. I'm all into grids - obviously, since I work for a grid company now - and i think it's the future, really.
    Hey Joe, if you are going to play up conspiracy theories about vendors, it's probably not a good idea to immediately start up a marketing pitch in the same post. Just a suggestion.
  26. Re: snake-oil[ Go to top ]

    The problem is that a lot of snake-oil salesmen on all sides are DESPERATELY invested in making sure their snake-oil sells - and they take any doubts as attacks. I'm all into grids - obviously, since I work for a grid company now - and i think it's the future, really.


    Hey Joe, if you are going to play up conspiracy theories about vendors, it's probably not a good idea to immediately start up a marketing pitch in the same post. Just a suggestion.
    No conspiracy theory here. Plus, if I was trying to be a snake-oil salesman, I'd have made a point to say how EVERYONE should use MY GRID, MY WAY, and no other product was ever necessary - and I probably wouldn't have pointed out that I work for a grid vendor now. I pointed out that I worked for a grid company because I have no interest in pretending I don't have a vested interest.
  27. Re: andrew[ Go to top ]

    I am literally holding my hands, trying not to type 'propaganda'
    Audible sigh, audible yawn. I typed a manual "Disclaimer" so the genesis of my opinions would be clear. And opinions, by definition, cannot be propaganda.
    a. EC2 is a total unknown, maybe some developers will rely on it for independent deployment, but will a single enterprise utilize for in-production deployment?
    I've already spoken with colleagues who are investigating the usage of EC2 as a disaster-recovery platform, and others looking to leverage it for scale-out. My personal survey is in no way scientific, but does indicate interest.
    b. It's not exciting, its a potential deployment alternative to Google, thats all...it will not pick up JPMorgan Chase, Caterpillar, or WalMart as anything more than testing environment customers, not their core systems...
    Thank you for your unresearched conjecture. I tend to follow the trends that the Enterprise will first investigate, then test, then pursue new opportunities; large mountains will not shift overnight. However, that my firm is offering bundled support specifically tailored for cloud computing is progressive as a business offering.
    c. JBoss 5 will be great, no one is really questioning that, and I am sure u r getting hard-core JBoss developers to do the modularization exercise by testing the pieces before the whole, but the impending release of JEE 6 and the day-by-day execution of SpringSource are quickly making it irrelevant in the marketplace for mass adoption - - i hate to even write those words, but that is the reality outside of Raleigh...
    The reality *inside* of Raleigh, where we actually measure things like Sales, Customer Satisfaction, etc, is that AS5 is in no way obsoleted. Despite a delayed release, attendance at the JBossAS 5 presentation @ JBossWorld was standing-room only.
    the only thing that should be going on right now in your organization is massive, coordinated, and expert dissemination of details of what to expect, what the ramp-up will be, what partners are signed on and testing betas, what dates to expect Seam 2.x, Exadel, and SOA integration, I mean just general marketing stuff...
    Strip the word "only" from the above, and I largely agree. This would be appropriate as a release approaches. S, ALR
  28. Somewhat "old" news...[ Go to top ]

    We have like half a dozen of our clients running GridGain on top of JBoss on Amazon EC2 in the last 9 months. I think it only underscores the fact that JBoss is a perfect platform to run applications on EC2. Apart of lack of IP-multicast on EC2 I don’t really see any significant quirks in setting up image for EC2 – and with GridGain you get ultimate flexibility in discovery and communication aspects of the grid. Best, Nikita Ivanov. GridGain - Grid Computing Made Simple
  29. Clustering JBoss in EC2[ Go to top ]

    Apart of lack of IP-multicast on EC2 I don’t really see any significant quirks in setting up image for EC2 – and with GridGain you get ultimate flexibility in discovery and communication aspects of the grid.
    I have not tried this, but ran across a relevant article regarding multicast/JGroups for EC2: http://www.hugotroche.com/my_weblog/2008/06/clustering-jbos.html S, ALR
  30. it IS expensive ,too expensive[ Go to top ]

    about $800 + bandwidth and storage...Why?!!!! what is the added value over normal ec2 to increase the cost 6 TIMES joe