Creating a SQL DataFlow

Intro

Create DataFlows to combine and transform your data using SQL queries. This is a more technical DataFlow solution than using Magic ETL but may provide a greater breadth of options. For information about creating a Magic ETL DataFlow, see Creating a Magic ETL DataFlow.

You access the interface for creating a SQL DataFlow from the Data Center. For more information about the Data Center, see Data Center Layout.

Video - MySQL DataFlow Overview

Video - Combining DataSets Using DataFlows

Creating a SQL DataFlow

To create a SQL DataFlow,

1. In Domo, click Data in the toolbar at the top of the screen.

2. Click SQL in the Magic Transform toolbar at the top of the window.

   Tip: You can also open the SQL DataFlow editor from anywhere in Domo by selecting  in the app toolbar and selecting Data > SQL.

3. Select the type of DataFlow you want to create.
   This opens the Create DataFlow view.

4. Enter a name and description for the DataFlow.

5. Select the input DataSets you want in the DataFlow by doing the following:

   1. Click Select DataSet.

      Note: Input DataSets in a DataFlow must already exist in Domo; you cannot upload new DataSets in the Create DataFlow view. For information about uploading new DataSets to Domo, see Connecting to Data with Connectors.

       

      Important: The name of the DataSet and each column's name in the DataSet MUST be 64 characters or less (including spaces.) This is a limitation of the MySQL engine but is not seen in Magic ETL.
   2. Select the desired DataSet.
      Information appears for the DataSet you selected, including the owner of the DataSet, details of its run history, column names, etc.
      
      

   3. Click Choose DataSet.
      A tile for the selected DataSet appears on the screen. You can click the DataSet tile to view the preview window.

      1. (Optional) Under the Select Columns tab, you can remove columns you don't want to include in the DataSet by clicking the "X" to the right of the column name, or remove all columns by clicking None.
         
         You can also add individual columns back into the DataSet by clicking Add Column and selecting the columns, or add all columns back into the DataSet by clicking All.
         
         

      2. (Select how the DataSet will be processed. Options include processing the Entire DataSet or Only new rows appended since the last DataFlow run.

         Note: This feature is currently in Beta. To be added to the Beta, please reach out to your Customer Success Manager (CSM).

          

         Important: You can only use the DataFlow append processing method if that input DataSet is configured to update with an append method. For more information on scheduling your DataSet with an append type update, see Basic Scheduling of a DataSet.
         Tip: If your input DataSet has a large number of rows, try using the append processing method. This will allow your DataFlow to run much quicker if it's only processing the new rows.

          

   4. Select as many additional input DataSets as you want by repeating step 5.

6. (Optional) Add transforms by doing the following:

   1. Click Add Transform.
      A Transform dialog appears with various options. For more detailed information about these options, see Understanding Transform and Output DataSet options.

   2. Enter SQL code to make the desired transformations to the input DataSet(s).

   3. Click Apply.

7. Click Add Output DataSet.
   An Output DataSet dialog appears with various options. For more detailed information about these options, see Understanding Transform and Output DataSet options.

   1. Enter SQL code to indicate how you want the input DataSets to be combined.
      If you have transformed the input DataSets in the Transform dialog, those transformations are applied here.

   2. (Optional) Add additional output DataSets by repeating the previous two steps.

   3. Click Done.

8. Schedule your DataFlow in the Settings pane. For more information on the different scheduling options, see Scheduling DataFlows.

   Note: If a schedule hasn't been specified, your DataFlow will default to a manual schedule.

9. (Conditional) If you want to run the DataFlow in Strict Mode, click the Settings button at the top of the screen then toggle the Strict Mode option to on.
   For more information about this option, see Understanding Strict Mode. 
   
   

10. (Conditional) Do one of the following to save your DataFlow:

    • If you want to run the script that outputs this DataFlow to a usable DataSet in Domo, click the orange down arrow in the upper right corner of the screen, select Save and Run, enter a version description if desired, then click Save to confirm.
      
      This starts the process of generating DataSets from the DataFlow. This generation process may take from a minute to an hour or more, depending on the size of the input DataSets. In addition to generating DataSets, a card for the DataFlow is added to the DataFlows listing in the Data Center.

    • If you want to save this DataFlow without outputting it to DataSets at this time, click Save, enter a version description if desired, then click Save again to confirm. A card for the DataFlow is added to the DataFlows listing in the Data Center, but no DataSets are generated. You can run the DataFlow to output DataSets at any time by mousing over the card for the DataFlow in the DataFlows listing, clicking , and selecting Run. This option and the other options available in this menu are discussed later in this topic.

When you save a DataFlow, an entry for this version is added to the Versions tab in the Details view for the DataFlow. If you entered a description when saving, that description is shown in the entry for the DataFlow. For more information about versions, see Viewing the Version History for a DataFlow. 

Understanding Transforms and Output DataSet options

When creating a new transform, you can choose to create either a table- or SQL-type transform.

• A Table Transform creates a new table using a SELECT statement and will always generate an output table. Due to an output table being generated, you can create easy indexes based on these tables. 

• A SQL Transform creates a table that typically doesn't include a SELECT statement such as a stored procedure. This type of transform does not generate an output table.

Both Transforms and output DataSets support the Run to here option. With Run to here, you can choose to run the DataFlow only up to the selected transform. In the case of very large DataFlows with numerous transforms or outputs, this may help you save time when testing your DataFlow's code. To use this option, select Run to here in the wrench menu for the transform or output DataSet, or open the editor for the transform or output DataSet, click the arrow next to the Run SQL menu, and select Run to here. 

You can also add transforms either above or below a given transform in your DataFlow. To do this, select Add transform above or Add transform below in the wrench menu for the transform, then select whether you want to create a table- or SQL-type transform. 

The following screenshot shows you the main components of the Transform dialog.
 

You can use the following table to learn more about these components:

Previewing the data

The preview will display in the transform after you have chosen Run SQL or Run to here. You can select options in the preview menu  to allow you to change how the results are displayed in the preview table for easier viewing.

The following components are available:

Easy indexing

An index is a data structure that improves the speed of operations in a table. With Easy Indexing, you can quickly add an index on one or more columns right to your input DataSet or table transform.

To create an index in an input DataSet or transform,

1. Open the editor for the input DataSet or transform.

2. Select the Indexing tab.
   
   

3. Choose your Index Type.

4. Select which column to apply the index.

5. Click Done. 

Video - SQL Explain Plans

Understanding Strict Mode

Strict Mode controls how MySQL handles invalid or missing values in data-change statements and also affects the handling of division by zero, zero dates, and zeroes in dates. You can find more info at https://dev.mysql.com/doc/refman/5.6/en/sql-mode.html#sql-mode-strict.

Best Practices for Creating DataFlows

Each DataFlow should...

• include descriptive names for each step of the transformation.

• include a description of the input DataSets being merged or manipulated and the DataSet being created, and should also indicate the owner of the data.

• be named the same as the output DataFlow—Because the outputs of a Dataflow become their own DataSet in the Data Center, this allows for easy identification of which DataSets are produced by which DataFlows.