Kettle 3 Milestone 2 is available

UPDATE: version 3.0 has been released!

Dear Kettle fans,

After a long period of bug-squashing and other frantic coding activities, we are happy to give you Kettle’s second milestone of version 3.0.0. (77MB zip file)

What has changed since M1?

  • New icons!! This is the first release to include a new set of icons and as such a fresh new look.
  • A new Mondrian Input step to read from Pentaho Analyses using MDX.
  • A new Regular Expression evaluation step
  • Access Input (don’t ask!)
  • Fixed / improved repository support
  • Improved database dialect handling (SQL Server .. problem and forcing identifiers to lower/uppercase)
  • Vast improvements to the clustering/partitioning engine (dynamic partitioning, etc)
  • A “Explore DB” button for quick access to your databases
  • A complete set of internal changes with respect to the “look & feel” of Spoon
  • A series of improvements to the Lazy Conversion algorithms
  • The remaining steps that where missing in M1 got ported
  • Of-course a host of bugs got fixed as well

With this 3.0.0M2 release we’re setting another big step towards more stability and maybe a GA release in a few months. However, please refrain from using this version in production situations!
Enjoy it!


Help us out and report bugs over here

11 thoughts on “Kettle 3 Milestone 2 is available”

  1. Cannot run PDI 3.0 M2: I have a bad error: “Could not find the main class. Program will exit”.
    I am running JDK
    Any idea ?


  2. Thanks for help: it works fine now.
    I installed it because I am looking at the remote execution of jobs: is this feature in 3.0 M2 ?
    Many thanks – Benoit

  3. Hi Matt,
    Is version 3.0.0 a beta version? And You said some more features will be available in 3.0.0 RC1 , sorry for ignorance but what does it mean by RC1?

  4. We don’t really work with alpha and beta releases.
    We typically release 2 or more Milestone releases (M1, M2, M3) in which new functionality is added.
    Then we release a series (usually 2 or 3) Release Candidates (RC1, RC2, RC3).
    You can call a release candidate a beta release if you like with that distinction that there can be several versions.

    It’s important to note that once we have a Release Candidate out we no longer add new functionality at all. Only bugs are fixed.
    Stability also (typically one must hope) goes up. In a week or so we’ll release RC2 which will be quite stable if I say so myself.

    Even though that is the case, we are still expanding the unit test-cases, enhance documentation, write new training material, do translations (i18n), make sure things are backward compatible, etc. Even though it’s totally different from purely coding Java, it’s still a lot of important work that needs to be done.

    After the Release Candidates we release a GA version which stands for General Availability. (targetted for the middle of November)

    Hope this helps,

  5. Hola Matt, que pena, hasta ahora estoy iniciando con esto de herramientas ETL, y pues la verdad en la mayoría de Foros recomiendan Kettle, por esto quisiera realizar la instalaci´ón y realizar unas pruebas, para saber si la empresa donde laboro, la utiliza para desarrollar un proyecto. Pero no he encontrado mucha información para realizar la instalación, estoy un poco confundido si para trabajar con Kettle debo instalar Mondrian..
    Gracias por tu ayuda.

  6. Hi José,

    There is absolutely no need to install Mondrian to use Kettle.

    The installation procedure for Kettle is very easy: download the Kettle binary distribution zip file and unzip it in a directory.
    Please note that you need version 1.5 of Java to run Kettle. So install a recent JVM first ( go to htt:// ).
    Then simply run “Spoon.bat” or “sh” if you are working on Windows or Linux/Unix/OSX.

    All the best,


Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.