Tag Archives: linux

Oracle buys Sun

Posted on by
Reply

via Between the Lines

As everybody already knows, Oracle is buying Sun for about $7.4 billion, including Sun’s debt ($9.50 a share in cash). Some interesting thoughts from Larry Dignan, Editor in Chief of ZDNet.

  • [Oracle] added that the acquisition of Java “is the most important software Oracle has ever acquired.”
  • Oracle also becomes a full-fledged hardware player.
  • Oracle and Sun have been long-time partners. […] “More Oracle databases run on the Solaris Sparc than any other system,” said Ellison, noting Linux was second. “We’ll engineer the Oracle database and Solaris operating system together. With Sun we can make all components of the IT stack integrated and work well.”
  • Oracle with Sun appears to be the Apple of the enterprise. Indeed, Oracle President Charles Phillips noted that the company is looking to offer everything from apps to the disk.
  • Oracle’s stack of IT stuff now includes:
    • Java;
    • Solaris;
    • Enterprise applications ranging from CRM to ERP to business intelligence;
    • The database (Oracle and MySQL);
    • The middleware;
    • The storage hardware;
    • Cloud computing services;
    • And servers.
  • Art of War approach:
    • Oracle gets to annoy IBM—and own Java—over a few pennies a share more than Big Blue was willing to pay.
    • Oracle gets to kill MySQL. There’s no way Ellison will let that open source database mess with the margins of his database. MySQL at best will wither from neglect. In any case, MySQL is MyToast.
    • Sun has a big installed base. All the better to upsell applications into.
  • Sun was relatively cheap compared to Oracle’s other acquisitions. The price was above the Hyperion buyout but below PeopleSoft and Siebel.
  • Oracle saves Sun management from what could have been a complete debacle following the IBM takeover talks. The Sun board had been split on the IBM deal. Today, it’s all roses.

The official Oracle press release can be found here.

It is clear that this deal will change the IT landscape quite massively. And the consolidation is definitely not finished…

Cross-posted on the Innoveo Blog.

Comparison of Open Source Application Servers

Posted on by
Reply

openlogic

I have found – by accident really! – a very interesting presentation on SlideShare concerning the different leading Open Source Application Servers on the market.

This very solid evaluation is proposed by OpenLogic.

I have summarized parts of it below.

Goal of the presentation /

Help you choose which open source application servers to evaluate more thoroughly:

  • JBoss
  • GlassFish
  • dm Server
  • Geronimo
  • Tomcat

Which open source application servers are your using or considering using /

appserver

J2EE Didn’t Fill the Need /

  • EJB 2 is difficult to use
    • Many people consider it unnecessary difficult
    • Entity Beans were especially problematic
    • But even session beans earned bad reputations
  • The Open Source community responds
    • OSS packages mitigating the problem
      • Hibernate instead of Entity Beans
      • Spring as a component model
    • They were very successful in the marketplace
  • Many Spring/Hibernate applications
    • Required only Tomcat to run but also work in full app servers

When Tomcat Is Not Enough /

  • Tomcat doesn’t support
    • JMS
    • EJBs
    • Other…
  • Enterprises need integration
    • The more mature the app is, the more likely it is to be added
  • Add-ons
    • It was possible to add things that Tomcat was missing
    • But this became a “build your own Frankenstein” exercise

Java Enterprise Edition in the Meantime /

  • Weaknesses in EJB 2 model were recognized
  • Work on new specification was completed
    • J2EE → JEE 5
    • EJB 2 → EJB 3
      • Entity Beans → JPA
  • EJB 3
    • Simple annotation-based programming model
    • Not everybody adopted EJB 3
  • JEE 6 will embrace modularity
    • Profiles, including web profile

Choices in 2007 /

  • Use Spring
    • Start with full app server
    • Start with Tomcat, build from there
  • Use EJB 3
    • Requires full app servers
  • Full application server
    • Pros – everything you are likely to need, could use EJB
    • Cons – complexity, might use more resources then Tomcat
  • Tomcat
    • Pros – small, simple, low resource usage
    • Cons – you might need to build your own app server, no EJB

OSGi /

  • Started in 1999
    • Recently got a lot of exposure
    • In particular, R 4.1
      • JSR-294
  • OSGi brings
    • Dependency management and modularity
    • Ability to load only parts it needs
  • OSGi currently has a lot of mindshare

The central questions /

  • Decisions, decisions
    • EJB 3 or Spring
    • Spring on dm Server or on J2EE/JEE server?
    • OSGi or not OSGi
    • Do I need EJB 2 compatibility?
    • In addition, there are many “old” considerations
  • Servers are not “all inclusive”
    • EJB 3 apps won’t work on SpringSource dm Server or Tomcat

Assessment JBoss /

JBoss Thought Leadership

  • Complete ecosystem
    • Portal
    • ESB
    • BPM
  • JBoss has history of innovation
    • Pioneer of EJB 3
    • Seam Application Framework
    • Web Beans
  • OSGi
    • Support in JBoss 5

JBoss for Developers

  • Excellent customizability
    • JMX-based, don’t deploy what you don’t need
  • Seam is worth a look for developers
  • JDK 6 with 4.2.3 and 5.0 GA
    • JDK 5 compiled binaries work on both JDK 5 and JDK 6
  • JBoss IDE
    • Eclipse-based
    • JBoss Tools
      • Free version (RHDS is paid version)

JBoss in Production

  • Dependability
    • Excellent clustering and failover capability
    • Reliable in production
  • Monitoring and deployment capabilities
    • Not really oriented toward system administration out of the box
      • Command line/file edit flavor of configuration
      • GUI tools (Tomcat manager and JMX Console) are fairly basic
    • Excellent 3rd party tools available for monitoring
      • Hyperic
      • GroundWork IT
      • JON *

JBoss – Conclusions

  • Strengths
    • Mature, scalable and reliable
    • Good support for J2EE and EJB 3
    • Seam framework
  • Weaknesses
    • Limited GUI-based configuration in open source version
    • LGPL License may be a concern for ISV’s embedding app servers

Assessment SpringSource dm Server /

SpringSource dm Server

  • Newcomer
    • Released in April of 2008
    • GPL license
    • Incorporates many mature components
      • Spring Framework
      • Tomcat
      • Equinox
    • dm Server users are very early adopters
  • Different take on app server
    • OSGi support
    • No support for EJB
    • No JMS out of the box

dm Server for Developers

  • Server works well with Spring Framework
  • IDE
    • Spring IDE – IDE for Eclipse platform
    • Support for NetBeans and IntelliJ
  • OSGi support
    • Resolves “dependency hell”
    • But requires application migration to take advantage of it
  • Supports Java 5 and 6

dm Server Thought Leadership

  • OSGi-based
    • Good OSGi implementation
    • OSGi discussed a lot in their documentation
  • No support for EJB
    • No support for EJB 3 or old EJB 2 spec
  • Spring offers similar functionality to Seam
    • Which is somewhat more mature
    • Although some of the Seam ideas might be somewhat more powerful
      • Bijection

dm Server – Conclusions

  • Strengths
    • Support for Spring Framework
    • Support for OSGi
  • Weaknesses
    • Newcomer
    • No EJB
    • Limited experience among workforce

Assessment Tomcat /

Tomcat

  • First release (3.0.x) in 1999
    • Apache license
  • Servlet container
    • Lightweight server
  • Used in many other app servers
    • JBoss
    • dm Server
    • Geronimo

Tomcat – Conclusions

  • Strengths
    • Lightweight
    • Well known and tested

      • < li>Fast startup/deployment for development

  • Weaknesses
    • No support for EJB, JMS or almost anything else outside of “web side”
    		•

    Tomcat in Development and Production

    • Tomcat 6 supports
      • Clustering
      • Failover
    • Widely used for both development and production
    • Supported in most popular IDEs

    Conclusions /

    • App servers are not just about the JEE specs
    • Make some high-level decisions before evaluation
      • Do I need EJB 2 compatibility?
      • Do I intend to follow EJB 3 and other industry standards?
      • Do I need something fast, lightweight, and easy to use?
      • Do I have a need for lifecycle management of server components (through OSGi)?
      • Do I need support for dynamic languages like Groovy and JRuby?
      • Am I an early adopter of new technology?

    Recommendations /

    • “I’m using EJBs and I’m conservative”
      • JBoss, GlassFish
    • “I don’t need XA/JMS/EJB”
      • Tomcat, JBoss, GlassFish, dm Server
    • “I’m using Spring”
      • Conservative – JBoss, Tomcat, GlassFish
      • Leading edge – dm Server
    • “I need to embed an app server in my commercial code”
      • Tomcat, Geronimo
    • “I use Spring heavily and I need OSGi”
      • I need it today and don’t need EJBs – dm Server
      • I need it soon and/or need EJBs – GlassFish, JBoss
    • “I use Seam”
      • JBoss
    • “I want ActiveMQ/Spring/Hibernate preinstalled”
      • Geronimo
    • “I need dynamic language support”
      • Groovy/Grails – dm Server, JBoss, GlassFish
      • JRuby/Rails – GlassFish, Geronimo

    Impressive figures from the Linux kernel development

    Posted on by
    Reply

    via 451 CAOS Theory

    Great post about some impressive figures from the Linux kernel development.

    • There have been almost 10,000 patches in each recent quarterly Linux kernel release.
    • Releases include work from ~1,000 developers and ~100 companies.
    • Since 2005, Linux has had more than 3,600 individual developers and more than 250 companies contributing to the kernel.
    • The individual development community has tripled in the last three years.
    • The top 10 developers have contributed 15% of changes, and the top 30 developers have contributed 30% of changes to the kernel.
    • Linus Torvalds is 27th on the list of contributors with most changes over the last few years. He has 495 to his name.
    • More than 11,500 or 14% of kernel changes have come from developers with no commercial entity backing their Linux development.
    • Another 13% of changes come from developers with ‘unknown’ commercial affiliation.
    • When we get to actual companies, Red Hat leads with 9,351 kernel changes, or 11.2%. Next is Novell with 8.9%, IBM with 8.3% and Intel with 4.1% of kernel changes.
    • More than 70% of all kernel development is demonstrably done by developers who are being paid for their work.
    • From the 2.6.11 kernel to the 2.6.24 release (1,140 days), there were an average of 2.8 accepted patches applied to the Linux kernel tree per hour.
    • An average of more than 3,600 lines of code is added to the Linux kernel tree every day.
    • Since 2005, the kernel has grown at a steady rate of 10% per year.

    Gosh, if this is not showing the incredible health and dynamism of the Linux community!

    cross-posted on the innoveo blog.