Data Connection Choices

Do I need to import my data into Tableau?

No. Tableau believes that you should be able to connect to your data in whatever way makes sense for you or your enterprise. You have two options with Tableau: connect live to your databases and files (leveraging the power you have already invested) or import your data into our lightning fast Hyper data engine. There are benefits to both approaches : our In-Memory or Live Data: Which Is Better? whitepaper does a great job outlining them. The biggest benefit is that with Tableau you get the choice. It is critical as a sales person you understand this point : it is probably our 2nd or 3rd biggest competitive differentiator. • Customer Consulting: this is a CRITICAL point for competing with Qlik! Our architecture aware data engine is still stomping them. This is a must-discuss item on Qlik deals.

Why does a live connect option matter? Are there benefits?

Connecting live can have important benefits such as: • It leverages the investment the customer has already made in their databases (e.g. security or speed). This avoids the time and expense of rebuilding this logic. • No scripting means they can connect and go with a few clicks - removing a big barrier for business users to start asking questions of their data • Their data stays vendor agnostic and open to all. When they import data typically only the vendor who imported it can then connect to it

What is the difference between 'Connect Live' and 'Extract'?

Connecting live is just as it sounds - you are establishing a live connection from Desktop to the original data source. When you drag-and-drop Tableau sends a query to the original data source leveraging its computational power and storage. When the data changes your live connection will reflect that once refreshed. When you use an extract you are taking a static snapshot of data from the original source and loading it into Hyper Tableau's own database technology. Now when you drag-and-drop the query goes to the Hyper extract protecting the original data source. Because it is a snapshot when the live data changes your view will not automatically reflect that. However you can put the extract on a refresh schedule (whole or incrementally append data) using Tableau Server (and Tableau Online depending on the data source).

When is best to use a live connection?

Good times to use a live connection can include: - If your database is already fast. Just connect and go! - If you need your data to refresh more often than every fifteen minutes we usually recommend a direct connection as well.

I don't see my data source listed in 'Connect To Data'? What are my options?

One option is to use the standard ODBC (Open Database Connectivity) option listed in the data connection window. While this connection option may not be as robust as a native connection it does provide the ability to use the driver for connecting to your database. If you run into performance issues I recommend extracting the data into a Tableau Data Extract. We can also provide some guidance to tune the connection. See Customizing and Tuning ODBC Connections You may also use Tableau SDK to write a program to access and process your data to create a Tableau Data Extract. We support developers on Windows and Linux platforms for Python/C/C++/Java. Another option for some customers is to build a Tableau Web Data Connector for data that lives on the web. Our Web Data Connector is an extensible framework to allow the Tableau community to connect to web based data sources.

How many data sources does Tableau connect to?

Tableau believes you should be able to connect to all the data you care about whether on-premise in the cloud big data like Hadoop or local files like Excel. We connect to several dozen data sources natively (plus many others through ODBC or Web Data Connector) and this list is growing all the time. The easiest way to see a current list is to open Tableau Desktop click 'Connect to Data ' and check out the current list right in the connection window.

What's the maximum number of rows Tableau can connect to at one time?

Tableau has no row limitation. Customers use Tableau to access petabytes of data because it only retrieves the rows and columns needed to answer your question. For example take a 100 million row database. Ask: what is the monthly sales trend for the last 2 years? Tableau retrieves 24 rows of data.

What is the Tableau Data Engine (powered by Hyper)?

The Data Engine is Tableau's high-performing analytics database for extracts on your PC. It compresses your data and uses columnar store : in some cases make the data faster to query. The Data Engine has the speed benefits of traditional in-memory solutions without the limitations that your data must fit in memory. And in Tableau's tradition of making powerful tools accessible to all there is no custom scripting needed to use the Data Engine.

What is an Extract?

This is Tableau's own file type Hyper (.hyper) which is a snapshot of the data stored on disk and loaded into memory as required to render a Tableau viz. An extract can be configured to pull a portion of data based on a condition : roll up granular data to a higher aggregation : or pre-filter data before bringing it into Tableau. See Extracts at Hyper Speed: What you Need to Know blog Use cases to NOT use an extract: • when live updates are needed • to replace a data warehouse

In a live connection what performance should I expect?

This is a great opportunity to do discovery first: What are your performance requirements? Is performance an issue with your current solution or solutions you're exploring? In a live connection performance speed is based on how fast your data sources can process requests made by Tableau. If live connection to a data source is too slow it may be possible to increase performance via extracting into Hyper.

Data story. What is our data philosophy?

We believe in letting people connect to the data they care about regardless of where it is. Data prep should be easy. Combining disparate data sources into a single source should be easy. And we make it easy for you to model and change your data.

Can I write my own SQL?

Yes Tableau does support custom queries. But first we recommend you try a native connection where Tableau generatesthe query and then measure the performance. Often it will be faster.

Can I use a table of data on a web page (e.g. Wikipedia or IMDB)?

Yes. Select the table and paste into a Tableau workbook via 'Data' in the toolbar.