Vault
NIM data layer
The Vault is the current collected view of your connected systems—the data NIM uses to make identity decisions and apply changes.
During collection, NIM reads the selected data from connected systems and loads it into the Vault. NIM workflows operate on that current Vault data, not on a live query to the external system.
How data moves through NIMDirect link to How data moves through NIM
| Stage | What happens | Where to configure it |
|---|---|---|
| Collect | NIM reads selected tables and columns from a connected system into the Vault. | Collection and data models |
| Interpret | NIM uses tables, keys, and relations to understand how collected records connect. | Data models and inter-system relations |
| Decide | Filters select and shape the current Vault data for a business outcome. | Filters |
| Act | Mappings and roles use filter output to make supported changes in target systems. | Mappings, roles, and jobs |
Current data versus collection historyDirect link to Current data versus collection history
| Data state | Available to filters and workflows? | Use it for |
|---|---|---|
| Current Vault data | Yes | Normal filters, mappings, roles, reports, and jobs. A successful collection replaces the current data for that system. |
| Past collection snapshot | No, until you load it | Comparing prior results, investigating data changes, or restoring a known snapshot to the Vault. |
NIM retains past collections on disk per system. Historical snapshots are inactive until you load a past system collection. Loading one changes the data that downstream workflows will use, so validate it before running a job.
View the current Vault dataDirect link to View the current Vault data
The Vault does not have a dedicated NIM Studio screen. To inspect it:
- Open a system table and use its Data tab to review the current collected rows.
- Create or edit a filter to see the current Vault tables available to your workflow.
- Review the latest collection log to confirm when the data was loaded and whether the collection reported warnings or errors.
See Collect and review system data for the complete collection and history workflow.
Keep Vault data dependableDirect link to Keep Vault data dependable
- Collect after changing a system connection, selected table, column, key, or relation.
- Use collection guards to stop unexpected source-data changes before they reach downstream workflows.
- Inspect a sample of table data after a material change, especially before enabling a scheduled job.
- Keep filters focused and use a consistent naming convention so their downstream purpose remains clear.
- Use dataset statistics when you need deeper validation metrics for an imported dataset.
NIM writes changes to external target systems when jobs run; it does not use mappings or roles to change the current Vault data. Run a new collection when you need the Vault to reflect the latest state of an external system.