Hitachi Vantara Pentaho Community Forums
Results 1 to 4 of 4

Thread: CloverTL comparison: CloverETL, Talend and PDI

  1. #1
    Join Date
    Jul 2007
    Posts
    247

    Default CloverETL comparison: CloverETL, Talend and PDI

    CloverETL released a comparison including a benchmark of CloverETL, Talend and PDI: http://www.cloveretl.org/_upload/clo...%20Pentaho.pdf

    Did anyone read this? From browsing the document, it looks like PDI didn't perform that well, especially with the large dataset. I'm afraid it's always a bit difficult to say if those tests are impartial or not, but at least they give a bit of an idea which parts of PDI are flawed or could be improved...

    I like the idea of Matt's comparison with Talend of sharing the test data with everybody so that one can perform his own benchmarks...

    What are your comments on the results?


    Regards,
    Ben
    Last edited by BeLienig; 03-09-2009 at 10:46 AM.

  2. #2
    Join Date
    May 2006
    Posts
    4,882

    Default

    Maybe make a virtual image of an environment, loaded with data and an instruction on what to extract, and then let the people experienced with a tool make the ETL

    E.g. for in the benchmark I wouldn't have used merge join... it's comparing apples and oranges

    Regards,
    Sven

  3. #3
    Join Date
    Nov 1999
    Posts
    9,729

    Default

    I agree with Sven,

    He could have just pasted the TCP-H query into the Table Input, but instead picked as many steps as possible to do work that the database should be doing. I'm sure there is a point to the benchmark, but it escapes me at the moment. Personally I have nothing against Clover ETL, it's yet another option for open source ETL, but this way they're making a bloody fool of themselves.

    Besides all that I found it interesting that the "author" of the PDF (can't call it a benchmark) also found out that Talend doesn't scale and runs in a single thread. If he would have taken the time to open a Sort dialog to tweak the settings a bit he would have found that Kettle would have kept running for all tests as well.

    The "author" then went trolling on the BOBJ forums as well in the hope that the Talend/PDI communities wouldn't kill him instantly:

    http://www.forumtopics.com/busobj/vi...c6fb9b3bfe4112

    In the end, he got slammed over there as well.

    'nough said ;-)

    Matt

  4. #4

    Default

    Quote Originally Posted by MattCasters View Post
    I agree with Sven,

    He could have just pasted the TCP-H query into the Table Input, but instead picked as many steps as possible to do work that the database should be doing. I'm sure there is a point to the benchmark, but it escapes me at the moment. Personally I have nothing against Clover ETL, it's yet another option for open source ETL, but this way they're making a bloody fool of themselves.

    Besides all that I found it interesting that the "author" of the PDF (can't call it a benchmark) also found out that Talend doesn't scale and runs in a single thread. If he would have taken the time to open a Sort dialog to tweak the settings a bit he would have found that Kettle would have kept running for all tests as well.

    The "author" then went trolling on the BOBJ forums as well in the hope that the Talend/PDI communities wouldn't kill him instantly:

    http://www.forumtopics.com/busobj/vi...c6fb9b3bfe4112

    In the end, he got slammed over there as well.

    'nough said ;-)

    Matt

    Having used CloverETL from its beginning, all I can say is that I am so glad I found Kettle.

    You have no idea!!
    Pentaho Data Integration CE 5.3.0.x
    JDK 1.7
    OS X Yosemite version 10.10.x
    MySQL 5.5.37
    Amazon Redshift
    Pacific Standard Time

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  
Privacy Policy | Legal Notices | Safe Harbor Privacy Policy

Copyright © 2005 - 2019 Hitachi Vantara Corporation. All Rights Reserved.