Denodo Tableau Exporter - User Manual

Download original document


What is the Tableau Exporter?

The Tableau exporter is a custom exporter designed to allow Denodo Scheduler Server to export data from different sources into a hyper file according to the version of Tableau.  

Hyper are data file format used by Tableau, a business intelligence software that helps people see and understand their data.

How to install the Tableau Exporter

If you're using Tableau 10.5 and later you have to install Extract API 2.0. See the How to install Extract API 2.0. Old versions don't support Extract API 2.0.

Additionally, you have to download Tableau Exporter from the Denodo Support Site. Once uncompressed, connect to the Scheduler Server via the Scheduler WebAdmin Tool and log in.

Once connected, click on Configuration → Plugins in order to access the Plugins Configuration section.

Click on the New Plugin button in order to access the new plugin input form.

Select the Tableau Exporter’s jar with dependencies file, which will be called something similar to denodo-tableau-exporter-[denodo_version]-[exporter_version]-jar-with-dependencies.jar (version numbers might change), and click on the Accept button, and after the success page, you will see your new exporter in the Plugins section.

How to install the Extract API 2.0

The Extract API 2.0 is available on the Tableau website. Go to the https://onlinehelp.tableau.com/current/api/extract_api/en-us/help.htm#Extract/extract_api_installing.htm%3FTocPath%3D_____3 page and choose the appropriate version for your platform. The programming language must be Java.

Note that the data extract process takes part in the Scheduler Server and therefore this is the server that must be configured.

Windows systems

  1. Go to the web site of the Tableau Extract API 2.0.
  2. In this page, sign in and download the package C/C++/Java 64-bit version for Windows
  3. Decompress the downloaded file.
  4. Add the bin folder to the System Variable PATH.

Linux systems

For Linux, follow these steps:

  1. Go to the web site of the Tableau Extract API 2.0.
  2. In this page, sign in and download the package C/C++/Java 64-bit version
  3. Decompress the downloaded file.
  4. Include in the System Variable PATH the folder /bin/hyper.
  5. Copy the content of the folder install-dir/lib[64]/tableausdk to $DENODO_HOME/dll/db.
  6. The libpcre16.so.0 package is a required dependency, if you are using a newer version of ubuntu, it is probably that you have to install this package and to create a symbolic link:

sudo apt-get install libpcre16-3

sudo ln -s /usr/lib/x86_64-linux-gnu/libpcre16.so.3 /usr/lib/x86_64-linux-gnu/libpcre16.so.0

  1. If the used distribution is CentOS or  OpenSuse, libcurses5 is a required dependency in the first case, and ncurses-compat-libs in the second one.

Install the library in CentOS:

sudo yum install ncurses-compat-libs

Install the library in OpenSuse:

sudo zypper install libncurses5

In this link it is also explained how to configure the installation to use Extract API 2.0. And in this link there is a section about troubleshooting with the Extract API.

Using the Tableau Exporter

The exporter offers some configuration parameters:

Mandatory

  • Output folder (String): It specifies the folder where the output files will be generated.

  • Output file name (String): Name of the generated file. The extension of the file has to be .tde or .hyper according to the version of tableau and it is mandatory. It offers support for placeholders that will be filled with execution parameters. The supported parameters are:

  • @{projectName}: name of the project
  • @{jobName}: name of the particular job.
  • @{jobID}: id of the particular job
  • @{jobTime}: time of execution of the job, in the format yyyyMMdd-HHmmss

NOTE: adding @{jobTime} to the output file name is normally a good idea to avoid file name collision, which might result in previous output files being deleted. For example, a useful output name could be: @{projectName}_@{jobName}_@{jobTime}.tde or @{projectName}_@{jobName}_@{jobTime}.hyper

Optional

  • Remove file first if it already exists (Boolean): If true the file specified will be removed if it already exists. Default: "false"
  • Append content if the file already exists (Boolean): If true the new content will be added to the existing file. Default: "false"

Creating a Tableau file

Select a job and edit it to add a new Tableau exporter.

Click on Edit Job → Exporters Section  New Exporter  TableauSchedulerExporter.

Fill, at least, the mandatory fields. For example:

  • Output folder: C:\Temp
  • Output file name: @{jobName}_output_file.tde

Execute the job, and the client_export_output_file.tde will be created into the selected folder. You can see the data by opening the file using Tableau Desktop and build views as you can see below:

Limitations

  • Tableau Exporter only was tested with the version of Extract API 2.0 2018-3-0.

  • The export to hyper files is only supported by Denodo Platform 7.0 and later.

  • Under certain circumstances, exporting Hyper files in Linux requires the server to be run by the root user. This is a requirement of certain versions of the Tableau Extract API 2.0. Please refer to your Tableau documentation or support for more detail