Configure the CSV directory integration settings

In the Admin Console, go to DirectoryDirectory Integrations.
Click the new CSV directory.
Click the Provisioning tab.
Select Integration in the Settings menu.
Complete these fields:
1. Connect to these agents: Select an agent. The agents must be version 01.03.00 or later. You can't use any previously installed Okta Provisioning Agent such as versions earlier than 01.03.00.
2. Unique identifier field: Enter a name that you want to use as the unique user identifier within the CSV file. For example, employeeid.
  - This should be the variable name for the attribute and not the display name.
  - This unique identifier is the sole identifying attribute for new or existing users and it must be enforced to be both unique and immutable.
  Once set, this attribute can't be changed.
3. CSV file path: Define the local file path and the .csv file name. This must be the absolute path, for example:
  - For Windows: C:\Users\Administrator\Desktop\csv\test.csv
  - For Linux: /opt/OktaProvisioningAgent/csv/test.csv
  For Linux, the .csv file must be located where the Okta On-Prem agent is installed. The default location is /opt/OktaProvisioningAgent/.
Optional. To set up incremental imports, complete these fields:
1. Connect to these agents: Select an agent. The agents must be version 01.03.00 or later. You can't use any previously installed Okta Provisioning Agent such as versions earlier than 01.03.00.
2. Unique identifier field: Enter a name that you want to use as the unique user identifier within the CSV file. For example, employeeid.
  - This should be the variable name for the attribute—not the display name.
  - This unique identifier is the sole identifying attribute for new or existing users and it must be enforced to be both unique and immutable.
  Note: Once set, this attribute can't be changed.
3. Deactivation field: Enter the field name that indicates user deactivation.
  During a full import, users who have the field marked and those who have been removed from the CSV file are deactivated. The deactivation field and deactivation value are optional, however both fields require configuration if you want to perform incremental updates. To apply these configurations to these attributes, the headers must be included in the CSV file.
4. Deactivation value: Enter a value that indicates when a user should be deactivated. Users with this value will be deactivated during an incremental import.
5. Full import filepath: Enter the file path and the .csv file name for full imports.
6. Incremental import filepath: Enter the incremental import local file path. This must be the absolute path and can't be the same as the full import file path.
If the incremental import file path is empty, a full update is performed during the scheduled update. If the incremental import file path isn't empty, an incremental update is performed instead.
Click Save.