After using ThoughtSpot DataFlow to establish a connection to an Apache Parquet file system, you can create automatic data updates, to seamlessly refresh your data.

To sync your data, perform these tasks:

Select file

  1. Immediately after creating a new connection, the connection detail page appears.

    You can fill out this information immediately, or return to it at a later stage, by clicking on the connection name in the list of connections.

  2. On the connection detail page, click Add table sync.

  3. In the Add table sync interface, on the database selector, scroll to find the source file, and click to select it.

    You can search for the file by name.

    If you use multiple files, you have to repeat these steps for each file.

  4. In the file detail display, adjust these properties:

    Column separator
    Specify the character that signals start of new column.
    The options are Comma(,) (default), Semicolon(;), Pipe(|), Space, Tab, or Other.
    Header row
    Specify if file uses a header row.
    Compression
    Specify if file uses compression.
    The options are None (default), or gzip.

    Before confirming that you plan to sync a file, examine it to ensure that it has the right information: the correct data types, and reasonable sample data.

    For wide files, scroll to the right to see all columns.

  5. Click Setup sync.

Specify sync schedule

Immediately after creating a new connection, the connection detail page appears.

When you schedule you create a sync, you have the following scheduling options:

  • Hourly
  • Daily(default)
  • Weekly
  • Monthly
  • Does not repeat

After you specify the sync schedule, click Save and sync now. This action saves the schedule, and starts the data sync immediately.

Alternatively, if you click Save, the system saves the schedule, but does not sync the data. This gives you the opportunity to fine-tune the column mapping between external data sources and tables inside ThoughtSpot, such as naming, visibility, data type conversion, conditions, and so on. The first sync starts on the specified schedule.

Hourly sync

Repeat
Select hourly.
Mandatory field.
Starts
Accept the defaults, or set your own starting date and time for syncing.
Note that the timezone is in UTC+00:00.
Mandatory fields.
Run the task every X hour(s)
Select the frequency of the sync.
The valid range is from 1 (default) to 12 hours.
Mandatory field.
Alert notifications
Specify when to notify of the sync status:
  • Failures (default)
  • Success and Failures
  • Never

Daily sync

Repeat
Select daily.
Mandatory field.
Starts
Accept the defaults, or set your own starting date and time for syncing.
Note that the timezone is in UTC+00:00.
Mandatory fields.
Alert notifications
Specify when to notify of the sync status:
  • Failures (default)
  • Success and Failures
  • Never

Weekly sync

Repeat
Select weekly.
Mandatory field.
Starts
Accept the defaults, or set your own starting date and time for syncing.
Note that the timezone is in UTC+00:00.
Mandatory fields.
Days of the week
Select the days of the week when you want to sync.
For example, you may select only Monday, or only Monday through Friday.
Mandatory field.
Alert notifications
Specify when to notify of the sync status:
  • Failures (default)
  • Success and Failures
  • Never

Monthly sync

Repeat
Select monthly.
Mandatory field.
Starts
Accept the defaults, or set your own starting date and time for syncing.
Note that the timezone is in UTC+00:00.
Mandatory fields.
Day of the Month
Select the days of the month when you want to sync.
Mandatory field.
There are two basic approaches:
Cardinal day
Click the Day selector, and choose by date of the month from drop down menu.
For example, select 15 to run sync on 15th of each month.
Ordinal day
Click the The selector. Then choose one of First(default), Second, Third, Fourth, or Last. Then choose one of the days of the week, Sunday through Saturday.
Alert notifications
Specify when to notify of the sync status:
  • Failures (default)
  • Success and Failures
  • Never

Does not repeat

Repeat
Select Does not repeat.
Mandatory field.
Alert notifications
Specify when to notify of the sync status:
  • Failures (default)
  • Success and Failures
  • Never

Map tables

To map the external files to ThoughtSpot’s internal database, follow these steps:

  1. Open the Advanced setup interface by clicking the toggle to open.

  2. Choose the Map tables and columns tab.

  3. Notice that the external database and table already appear, under External data source.

  4. Specify the following information for internal ThoughtSpot storage:

    Filename
    The name and path of the file
    ThoughtSpot database
    Select an existing ThoughtSpot database from the drop-down menu.
    If you wish to create a new database in ThoughtSpot, click TQL Editor, enter the CREATE DATABASE my_database; command, and click Execute.
    Mandatory field.
    ThoughtSpot schema
    Select an existing ThoughtSpot schema from the drop-down menu.
    If you wish to create a new schema in your ThoughtSpot database, click TQL Editor, enter the CREATE SCHEMA my_schema; command, and click Execute.
    Mandatory field.
    New or Existing table
    Mandatory field.
    Create a new table
    Choose this option when you want to load data into a new table.
    The system creates a new table automatically.
    If you want the new table to have some special properties, create the table using the TQL Editor, proceed to Choose existing table, and then select the table you just created. Click TQL Editor, enter the CREATE TABLE my_table ...; command, and click Execute.
    Specify the table name.
    Choose existing table
    Choose this option to load data into a table that already exists inside ThoughtSpot.
    Select the table name from the drop-down menu.
    ThoughtSpot table
    The name of the target table for data sync, inside ThoughtSpot.
    Mandatory field.

Map columns

To map the columns of the external file columns to columns in ThoughtSpot’s internal tables, follow these steps:

  1. Open the Advanced setup interface by clicking the toggle to open.

  2. Choose the Map tables and columns tab.

  3. Scroll down, to Map the columns… section.

  4. Specify the following information for columns:

    1. Adjust the file parsing properties as necessary: Column separator, Header row, and Compression.
    2. Select (or deselect) columns for syncing into ThoughtSpot.
      By default, all columns are selected.
    3. Search for columns by name; this is very useful for very wide tables.
    4. Rename columns in the ThoughtSpot table, for easier search.
      To make this change, click the pencil (edit) icon next to the name of the column.
    5. Change the data type of the column inside the ThoughtSpot table.
      Click the down chevron icon to open the drop-down menu, and select a new data type.
      For example, if you know you have integer data, change the default DOUBLE datatype to INT32.
    6. Set Primary keys of the table by toggling the selector to the ‘on’ position.
      Note that several columns may be primary keys.
    7. Set the Sharding keys of the table by toggling the selector to the ‘on’ position.
    8. Specify the Number of shards in the table.
  5. Save your work by clicking Save.
    Alternatively, click Save and sync now to save your work and sync data at the same time.

Set sync properties

Basic sync properties

To set the sync properties, follow these steps:

  1. Open the Advanced setup interface by clicking the toggle to open.

  2. Choose the Sync properties tab.

  3. Under Set sync properties, specify the following information:

    Sync mode
    Choose the sync mode, either Append or Overwrite.
    Mandatory field.
    Append
    This option adds new rows to the table.
    Overwrite
    This option removes all existing rows, and then adds new rows to the table.
    Completion
    Specify what to do with the file after the sync successfully completes:
    • Do nothing to the file
    • Delete the file
    • Move it to: directory/path
  4. Save your work by clicking Save.
    Alternatively, click Save and sync now to save your work and sync data at the same time.

Scripting for sync

To use scripting before or after the sync, follow these steps:

  1. Open the Advanced setup interface by clicking the toggle to open.

  2. Choose the Sync properties tab.

  3. Scroll down to Advanced configuration, and click the toggle to expand.

  4. Specify the following information:

    Wait for file
    Specify the fully-qualified name of the file (directory path and file name) that must be present before sync can start.
    Pre-script
    Add the script that you want to run before syncing.
    Optional field.
    For example, to drop data that is over 1 year old, enter the following expression:
    DELETE from FACT_TABLE Where DATE < ADD_DAYS(TODAY(),-365).
    To find the correct functions for the script, click Expression editor, use it to create a valid command (or several commands), and click OK.
    Post-script
    Add the script that you want to run after syncing.
    Optional field.
    For example, to drop data that is over 1 year old, enter the following expression:
    DELETE from FACT_TABLE Where DATE < ADD_DAYS(TODAY(),-365).
    To find the correct functions for the script, click Expression editor, use it to create a valid command (or several commands), and click OK.
  5. Save your work by clicking Save.
    Alternatively, click Save and sync now to save your work and sync data at the same time.

Connector-specific sync properties

To modify connection-specific properties for sync, follow these steps:

  1. Open the Advanced setup interface by clicking the toggle to open.

  2. Choose the Sync properties tab.

  3. Scroll down to Sync connector properties, and click the toggle to expand.

  4. Specify the sync properties for files:

    • Enclosing character
      Specify if the text columns in the source data needs to be enclosed in quotes.
    • Escape character
      Specify escape characer if data uses text qualifier is mentioned. This should be the character which escapes the text qualifier character in the source data.
    • Null value
      Specifies the string literal indicates the null value for a column. During the data load, the column value matching this string will be loaded as null in the target.
    • Date style
      Specifies how to interpret the date format
    • Date delimiter
      Specifies the separator used in the date format.
    • Time style
      Specifies the format of the time portion in the data.
    • Time delimiter
      Specifies the character used as separate the time components.
    • Skip trailer rows
      Skip trailer row while processing the data
    • Reader encoder
      Specify the encoding format of the source system.
    • TS load options
      Specify additional parameters passed with the tsload command. The format for these parameters is:
      --<param_1_name> <optional_param_1_value>
    • Boolean representation
      Specifies the representation of data in the boolean field

    See Sync properties for details, defaults, and examples.

  5. Save your work by clicking Save.
    Alternatively, click Save and sync now to save your work and sync data at the same time.