Collection
Collection guide
Collect current system data into the NIM Vault, reload prior snapshots, and safeguard collections against unexpected changes.
Work with system data
Collect
Query a system for its current dataset and load the result into the Vault.
Collect and load a system →Load a previous collection
Return a stored snapshot to the Vault without collecting new source data.
Load a past collection →Automate collection
Keep source data current by including collections in scheduled sync tasks.
Explore sync tasks →NIM interprets data in the Vault through the system's data model, making it available to filters, mappings, and other NIM workflows.
Collect versus load
| Operation | What it does | When to use it |
|---|---|---|
| Collect and load | Requests current data from the connected system, then stores and loads the result into the Vault. | Normal day-to-day operation and after configuration changes. |
| Load a past collection | Loads a previously stored snapshot into the Vault without making a new source-system request. | Roll back from invalid or corrupted data, or investigate when a change occurred. |
Every collection creates a timestamped folder under C:\ProgramData\Tools4ever\NIM\sysdata\systems\<system>\<YYYYMMDDHHMMSS>. NIM retains these snapshots indefinitely, so they can be reloaded when needed. Use past collection history to review the available snapshots.
Configure collection guards
Collection guards protect data quality by stopping imports with unexpected changes before they affect downstream processes. They are especially useful when a source system returns incomplete data, a large unexpected row change, or a changed column structure.
Set a baseline
Open the system, select Guards, enable the [default] guard, and set the expected row-count or percentage-change threshold.
Protect critical tables
Create a table-level guard when a table needs a threshold that differs from the default. A table-level guard overrides the default for that table.
Watch column changes
Use default or table-level guards to detect unexpected added or removed columns—especially for CSV imports.
Review and tune
Monitor guard results and adjust thresholds as you learn the normal variation in each system's data.