Hitachi Vantara Pentaho Community Forums
Results 1 to 4 of 4

Thread: Create OLAP cubes

  1. #1
    Join Date
    Jun 2016

    Default Create OLAP cubes

    Maybe someone can help...
    I've never tired to create OLAP cubes. This subject is absolutely new for me.
    I see Pentaho offers some products to create such data layer; Schema Workbench, Mondrian server...

    Suppose I have MySQL database. What I need to convert data into OLAP cubes, what layers I should organize with what tools? Pentaho Server is the only tool to display those data?
    Can I use Pentaho Kettle to do some part of the job?

    Any help appreciated,


  2. #2
    Join Date
    Jul 2009


    I used Mondrian and MDX years ago, and am looking into them again for a project I'm working on now. Here's my understanding. Others might have better explanations:

    Mondrian builds OLAP cubes in memory using data from a relational database such as MySQL or Postgres. The Mondrian schema file, which is just an XML file that you create, tells Mondrian where to find the dimensions and facts for MDX queries. Mondrian caches its results, so if it receives two MDX queries that use some of the same data, the second query will use the applicable cached results from the first query. When you load more data with ETL, you must tell Mondrian to clear the cache, or parts of the cache, and it will then re-pull data from the relational database as needed for future queries.

    Conceptually, Mondrian works kind of like a group of database materialized views.

    The Schema Workbench tool is for creating a Mondrian schema, which is just an XML file. You could create the file by hand if desired. I believe that Schema Workbench (free), and Pentaho's front end user tools like Pentaho Analysis Services ($$) and JPivot (free) only work with Mondrian 3. (You can also create Pentaho Report Designer reports that use Mondrian as their data source.) There is a Mondrian 4 that was released a few years ago, but it seems to be dormant.

    There are other front end user tools like Saiku (free or $$) and FlexMonster ($$) that can connect to Mondrian to issue MDX queries.

    These front end tools use a protocol called XMLA, or XML for Analysis, to communicate with a Mondrian server, and you could write your own code that issues XMLA calls if you wanted.

    Hope this helps.

  3. #3
    Join Date
    Apr 2008


    Much like the PRD can help you convert a DB Table into a formatted PDF, Mondrian can help you convert a structured DB to Statistical "Slice&Dice"

    OLTP(DB) -> PDI -> OLAP(DB) -> Mondrian -> PresentationLayer

    PresentationLayer can be PRD, Excel (with a plug-in), Saiku, etc.

    Much like we recommend using Spoon to design your transforms, and then Pan/Kitchen to run them, Schema Designer (PSW) helps you design what Mondrian does. There's also Aggregation Designer (PAD) that helps you pre-compile the data for Mondrian.

    Pentaho Server includes Mondrian, PDI, Pentaho Reporting, and a few other things. You *can* run Mondrian on its own, but it might not be worth the increased effort compared to using Pentaho Server.

  4. #4
    Join Date
    Jun 2016


    Thank you very much indeed guys for explanations.
    I tried Pentaho BI Server several times yet every time I had to give up. This tool never seemed interesting to me. There is something missing there; ergonomics, the possibility of data presentation ... for instance Kibana is way better tool. It's a pity that Pentaho does not offer good BI platform.I really like working with PDI and Report Designer. But I guess PRD will never be able to make clickable reports based on OLAP cubes? It can only produce static reports based on them? Which means that the use of data between the data and the OLAP report is unjustified.

Tags for this Thread

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.