> ## Documentation Index
> Fetch the complete documentation index at: https://help.dingtalk.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Data Factory FAQ

> Common questions about YiDA Data Factory, including database table name support, Data Source error fixes, firewall configuration, and cross-app data access.

<AccordionGroup>
  <Accordion id="q1" title="Q: Do database table names support both Chinese and English?">
    * Only English table names are supported.
  </Accordion>

  <Accordion id="q2" title="Q: How do I resolve Data Source errors (partial list)?">
    | Error message                                                                                                                                            | Solution                                                                                                                 |
    | -------------------------------------------------------------------------------------------------------------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------ |
    | data source validate failed.connectionInfo isn't correct                                                                                                 | YiDA only supports the public network. If the database does not expose a public network endpoint, YiDA cannot access it. |
    | ORA-00972: identifier is too long                                                                                                                        | The field alias is too long. For Oracle databases, keep the alias within 64 characters.                                  |
    | When the Data Source retrieves SQL Server data and the report uses aggregation, the error "concat ws is not a recognized built-in function name" appears | To use concat\_ws, upgrade to SQL Server 2017 or later.                                                                  |
  </Accordion>
</AccordionGroup>

### Q: When Creating a Data Source, the Error "Connection Test Timed Out. Check the Connection Information and Firewall, Then Try Again" Appears

* When integrating a Data Source with your own system, ensure your database or system is accessible over the public network. Otherwise, this error occurs.

### Q: When Using SQL Server (Versions Earlier Than 2012), the Error "Xxx Field Is Not a Recognized Built-In Function Name" Appears

* YiDA is not currently compatible with SQL Server versions earlier than 2012. Data may occasionally display correctly, but these versions are not recommended.

### Q: Trigger Rules Between Form Submission, Bulk Import (Trigger on Selection), OPEN API / Page Data Source, and Business Association Rules or Integration & Automation

| Action                             | Can trigger business association rules | Can trigger Integration & Automation |
| ---------------------------------- | -------------------------------------- | ------------------------------------ |
| Form submission                    | ✅                                      | ✅                                    |
| Bulk import (trigger on selection) | ✅                                      | ✅                                    |
| Open API / page Data Source        | ✅                                      | ✅                                    |
| Business rule                      | ❌                                      | ❌                                    |
| Integration & Automation           | ❌                                      | ❌                                    |

<AccordionGroup>
  <Accordion id="q3" title="Q: Are database view connections supported?">
    * Not supported.
  </Accordion>

  <Accordion id="q4" title="Q: How do I configure the IP firewall allowlist?">
    Use the following allowlist configuration for the network of YiDA external Data Sources and business services.

    | If your organization's database uses a public network IP allowlist,<br />add the following entries to avoid being blocked by firewalls or other network settings:<br />• 106.11.0.0/16<br />• 203.119.128.0/17<br />• 59.82.0.0/16<br />• 140.205.0.0/16 | Failed to connect to a database-type Data Source, with a prompt to check the firewall. |
    | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | -------------------------------------------------------------------------------------- |
  </Accordion>

  <Accordion id="q5" title="Q: Can subform data be retrieved across apps?">
    Yes. A subform can serve as an independent dataset and be selected for cross-app data retrieval.
  </Accordion>

  <Accordion id="q6" title="Q: In which scenarios can cross-app data be used?">
    Cross-app data is currently supported only in report scenarios. Form data linkage, multi-table association, and other scenarios will be supported gradually in the future.
  </Accordion>
</AccordionGroup>

### Q: When Will the Attendance Dataset Be Available?

A: It is on the roadmap and will be considered for delivery in future iterations.

### Q: Why Can't I Preview the Latest Data in the Official Dataset I Connected To?

A: You currently need to manually click the test button to refresh the data. A refresh button for preview data will be added later.

### Q: Is the Data Real-Time or Updated the Next Day?

A: Currently, only \[OA Approval] data is real-time. General metrics, employee roster, Smart HR (onboarding, transfer, promotion, resignation), and similar datasets are updated by default at 00:00 the next day.

### Q: After an Admin with Data Permissions for a First-Party App Creates a Dataset via Data Factory, How Can Data Permissions Be Further Configured?

A: Data processed through Data Preparation generates a new independent dataset based on the data permissions of the most recent publisher. The default row and column permissions of this dataset depend on the last publisher's permissions. On this basis, you can configure independent dataset permissions.

<AccordionGroup>
  <Accordion id="q7" title="Q: What is the data range of OA Approval? Does it include all OA Approval data?">
    A: Only OA Approval data generated within the past year is currently available.
  </Accordion>

  <Accordion id="q8" title="Q: Are there precision issues when using division in Data Preparation?">
    Floating-point conversion is currently required. A floating-point formula function will be provided in the future.
  </Accordion>

  <Accordion id="q9" title="Q: The dataset does not update after adding new components?">
    After adding new components, click test again or drag the dataset in again to update it.
  </Accordion>

  <Accordion id="q10" title="Q: The Data Preparation node preview result does not match the local calculate result?">
    The number of data rows shown in the data preview within Data Preparation is for reference only. It cannot be used to judge the final output dataset row count, because intermediate acceleration caching may exist.
  </Accordion>

  <Accordion id="q11" title="Q: Under what conditions does modifying the field type of a Data Preparation node take effect?">
    For example, if the nodes are structured as A - B - C, modifying the field type of node A affects only nodes B and C. Modifying the field type of node B affects only node C.
  </Accordion>

  <Accordion id="q12" title="Q: Why do the field IDs of output nodes change?">
    In daily use, avoid irreversible delete or reset actions on data operation nodes.

    * If the content of a preceding processing node changes (for example, it contains a field A with field ID = xxx001), including deleting and re-adding the node, a completely new field ID is dynamically generated for subsequent fields (the ID of field A may be randomly regenerated as yyy789, for illustration). This causes existing node ID configurations in downstream business consumption scenarios such as reports and dashboards to fail.
  </Accordion>

  <Accordion id="q13" title="Q: When Data Preparation retrieves DingTalk attendance data, the 'first check-in time of the day' shows null for some Users even though they do have check-in records. Why?">
    In YiDA, the exposed field "first check-in time" uses the format: `yyyy-MM-dd HH:mm:ss`.

    Display rules:

    * If the user has a shift schedule and clocks in for the first work shift, the time is recorded as the first check-in time of the day.
    * For Approval forms, out-of-office, leave, supplementary check-in, and field work records are not counted.
    * For check-in records, field work and admin-modified records are not counted.
  </Accordion>
</AccordionGroup>