Single Table Update of Self-Service Datasets

Version

Application Scenario

This document introduces the update settings of self-service datasets.

Note:

Self-service datasets refer to tables published from the data editing page to Data Directory. For data table updates within an analysis subject, refer to Editing an Analysis Subject.

Function Description

• Self-service dataset updates support Update Now or Scheduled Update.

• You can determine whether to update a table with its parent table.

• The simplest table does not need to be updated.

• The system will not update inactive tables to save update resources.

Self-service datasets with the following features are defined as simplest tables:

• Self-service datasets without associations;

• Self-service datasets with only field settings (without operations such as filtering, summary by group, or adding columns).

Note:

Self-service datasets based on Excel files uploaded directly to My Analysis are not simplest tables.

A simplest table retrieves data from its parent table instantly (requiring no separate data extraction or updates), which saves disk space. Select a simplest table and click Update Information. The message "No data needs to be extracted for this table. Latest data is obtained automatically.” is displayed in the system.

Update Now

Except for simplest tables, other self-service datasets in extraction status need to be updated to obtain the latest data.

Select the data table to be updated and choose Update Information > Update Now, as shown in the following figure.

Updating with the Parent Table or Not

Click Update Setting and select Update with Parent Table or Not Update with Parent Table, as shown in the following figure.

For example, the Data Lineage tab page of the Load risk analysis self-service dataset shows that this table has three parent tables.

• When you select Update with Parent Table, data in the self-service dataset is updated simultaneously as one of the three parent tables is updated.

• When you select Not Update with Parent Table, data in the self-service dataset is not updated simultaneously even when the three parent tables are updated.

If a parent table has multiple child tables, you can set Update Setting for some less frequently used ones to Not Update with Parent Table. In this way, less data can be updated, and the update time can be saved during the parent table update. For child tables with Update Setting set to Not Update with Parent Table, you can update them regularly through Scheduled Update.

Note:

If you separately update a child table whose parent table is a self-service dataset, its parent table will be updated as well.

Default Update Settings by the Admin

You (the admin) can determine whether to update self-service datasets with their parent tables by default under System Management > System Setting > BI Parameter, as shown in the following figure.

The modification of Default Update Settings for Self-service Dataset only applies to the self-service datasets created after the modification and does not affect those created before.

Scheduled Update

You can set scheduled updates in the Update Setting page.

• You cannot configure scheduled updates for datasets with the Update Setting set to Update with Parent Table.

• You can configure scheduled updates for datasets with the Update Setting set to Not Update with Parent Table.

For details about the configuration steps, see the "Scheduled Update" section in Single Table Update of Base Tables.

If an inactive data table is rarely used, it is not included in any updates to save update resources.

Click an inactive table. The message "Data is not active and will not be extracted currently. If you need to activate data, you can update it immediately in Update Information" is displayed.

If you need to use this table, click Go Now or choose Update Information > Update Now to update it. The following figure shows the settings.

If you view components or dashboards that contain inactive tables, the message "Edit Failure. Data is not active, which will be automatically activated after 5s.” is displayed. The data table is automatically updated the moment you view it. (The update is automatically triggered after about 5 seconds.)

On the mobile terminal, starting from V6.0.14, when inactive resources are viewed, they are also automatically updated after 5 seconds, as shown in the following figure.

Setting the Active Period of Data Tables by the Admin

You (the admin) can set the active period of data tables under System Management > System Setting > BI Parameter.

1. If Active Period of Accessed Dataset (Day) is set to X, a table is defined as active in either of the following cases:

• The dashboards/components made from this data table are edited or accessed within X days.

• The dashboards made from this data table are exported via the BI dashboard export API within X days.

• The child tables of this data table are edited and saved within X days.

In any of the above cases, this data table is defined as active.

2. If Active Period of Created Dataset (Day) is set to X, a table is defined as active within X days of its creation.

3. If Active Period of Created Dataset (Day) is set to X, a table is defined as active within X days of its creation.

4. After Global Check Time of Dataset Activity is set, new inactive tables can be found through the global check. You can customize the global check time.

For details, see BI/Spider Parameter.

Special Scenario: Inactive Table Activation Mechanism for Scheduled Tasks

Although dashboards pushed by scheduled tasks can be included in the activity logic, if the dashboard contains inactive tables internally, the first push will still fail. The system will automatically attempt to activate the table, but it does not take effect immediately. The following conditions must be met sequentially:

• The first push will inevitably fail: Only when the first push fails will the system trigger the action to "activate" the table (refreshing its preview time).

• Midnight of the next day must be reached: The system performs a global check only at a fixed time (default value: 00:00:00) each day. Only after this time point will the table truly change to "active status".

• A data update must be triggered (key step): After the table status changes to active, you must manually perform a single table update or use a scheduled task to successfully complete one data extraction update for this table.

• After you have completed the above steps, the scheduled task can be executed normally.