Tableau Bridge Extracts: Centralizing and parallelizing

Version 1

    Tableau Bridge lets publishers refresh published extracts directly from their computer, with little overhead. This means letting publishers refresh published extracts directly from their computer, and may mean many Bridge clients in your organization. However, not all organizations want to manage Bridge clients on a multitude of computers. Centralizing your extract refreshes on a single Bridge client is one approach to addressing this. Another alternative is to run multiple centralized clients so you can schedule extracts in parallel. Here's a quick guide to both of these approaches.

     

    Part 1: Centralizing Extracts

    Here's a short recipe (sorry, I like to cook ) that will allow publishers to publish their extract data sources and then transfer the refresh schedule to a central client.

     

    Before you begin, you’ll need:

    • One publisher
    • One extracted data source
    • One site administrator
    • One or more Tableau Bridge clients

     

    Steps for publishers:

    1. Publish your data source to Online as an extract. You do not need to associate it with a Bridge client, but it can be.

    2. Ask a Online site administrator to "own" the refreshes of the data source (they can use the steps below).

    • Be sure to give them the name of the datasource, and a schedule on which you want it refreshed.

     

    Steps for site administrators:

    1. Sign in to Tableau Online, find the published data source, and change ownership from the publisher to yourself (assuming you're signed in as the site admin). picture 1snap1.png

    2. In the Online Web UI move to the Data Source page and select the Refresh Schedules tab.

    3. Next to “Refresh this extract on:” (on the far right side of the window), select the centrally-managed Tableau Bridge client. (See the picture below)

    4. Set the schedule the publisher asked for.

    Pickit2.png

    picture 2

     

    The new extract schedule will appear in the central Tableau Bridge client shortly. Future refreshes will be completed on this client.

     

    More approaches to delegating refreshes

    You can get a bit fancier with how you delegate refresh schedules to clients. The following recipe demonstrates two more advanced techniques. Delegating Bridge client management to a publisher account (this lets you move management away from a site admin account, which you may want to do if you're moving management away from the Site Admin role) and allowing the Bridge Client Owner to pick one of several clients to run the refresh on (so you could potentially refresh extracts in parallel).

     

    Before you begin, you’ll need:

    • One Published extracted data sources (or more)
    • One site administrator account
    • One publisher account
    • Two or more Tableau Bridge clients

     

    1. Change ownership of the data sources to a centralized publisher account (if this isn't already done):

    • In Tableau Online, sign in as Site Administrator.
    • Navigate to the Data Sources tab.
    • Select the … next to your data source and select Change Owner from the … menu. (see picture 1 above)
    • Change the owner to the publisher account.

    2. Set up two separate machines (typically virtual machines) with Tableau Bridge clients installed.

    3. Sign in to the Tableau Bridge clients to Tableau Online using the centralized publisher account.

    4. Select the Data Sources tab in Tableau Online (you can get there from the Tableau Bridge client by clicking the site name link).

    5. Next to “Refresh this data source on:” (on the far-right side of the window), select the Tableau Bridge client you’d prefer and adjust the extract schedule. (see picture 2 above)

     

    Note: If your data source requires username and password authentication, you will need to enter it on the new Tableau Bridge client. You should also make sure the virtual machine in question has access to the underlying data. You can use the Test Connection button in the Bridge client to check.

     

    Why can’t a publisher assign a data source to a centrally-managed client?

    Tableau Bridge client owners can only assign data sources to the clients that they “own” (or are signed into). A site administrator can become the owner of any data source, which allows them to take a refresh schedule from one client and move it to one they “own.” There are several reasons for enforcing this kind of ownership model:

    1. Publishers aren't always aware of the status of computers besides the ones they “own,” so allowing publishers to refresh extracts on other Tableau Bridge clients can lead to refresh failures.
    2. The Tableau Bridge client needs access to the underlying data source. That is most likely to be the machine the Publisher published from, not some other machine.