Data tables
Data tables guide
A data table represents one kind of record from a connected system—such as employees, users, groups, or memberships.
Data tables are the foundation of a system’s data model. During collection, NIM reads the selected tables and columns, then stores their rows in the Vault for use in filters, mappings, and roles.
Tables, columns, and rowsDirect link to Tables, columns, and rows
| Part of the model | Represents | Example |
|---|---|---|
| Table | One object type exposed by a connector. | employees, groups, or memberships |
| Column | One attribute of that object type. | employee_id, email, or department |
| Row | One individual record in the table. | One employee, group, or membership record |
For example, an Active Directory system can expose tables for users, groups, group memberships, and organizational units. A row in the users table is one directory account; its mail column contains that account’s email address.
Choose the data your workflow needsDirect link to Choose the data your workflow needs
Start with the business outcome, then collect only the data required to reach it.
| If you need to… | Usually collect… |
|---|---|
| Provision accounts from HR data | Employee records, a stable employee identifier, name, employment status, department, and manager information. |
| Assign access by organization | Employee records plus departments, job titles, locations, or other organizational reference tables. |
| Manage group membership | Users, groups, and the membership table that connects them. |
| Match accounts across systems | A reliable identifier in each system, such as employee ID, immutable ID, or directory object ID. |
Do not treat display names or email addresses as universal identifiers. They can change or be duplicated. Configure a key from a stable, unique attribute before using a table in relationships or matching logic.
Configure tables and columnsDirect link to Configure tables and columns
- Choose tables to collect from the system’s available tables.
- Choose columns to collect for each selected table, and set their appropriate data types.
- Assign primary keys to the tables that need stable record identity.
- Add intra-system relations when one table refers to records in another table in the same system.
- Collect the system and review the collected data before creating filters or mappings.
Set a type for every collected column. Incorrect or missing types can cause filter errors or produce unexpected comparisons.
Add calculated columns when source data needs shapingDirect link to Add calculated columns when source data needs shaping
Use a custom JavaScript column when the value you need can be safely derived from collected fields—for example, combining a first and last name or normalizing a source value. Keep the original source columns as well when they may be needed for troubleshooting or future workflows.
Custom JavaScript in this area is part of a system’s collected data model. It is different from NIM App scripts, which run only in Apps.