Importing data

You rarely start from nothing. Layertrail imports from two sources, and both show a full preview before anything changes. The preview and the import use the same planning code, so what you see is exactly what you get. An import undoes in one step.

CSV

Choose File → Import… (Cmd/Ctrl+I) and pick a CSV file. To import under a specific node, right-click it and choose Import rows here…. Layertrail reads the header row and a sample, then asks how to map it:

• Name column. Which column becomes each node’s name. If names repeat (common in asset lists), Layertrail suggests a column with unique values instead.
• Placement.
  • Flat: every row becomes a child of the node you choose.
  • Path: a breadcrumb column such as Lab B / Rack 4 says where each row goes, walking down from a base node. Missing parents can be created automatically.
• Property columns. Include or drop each remaining column and choose its property key.
• Template. Optionally bind imported nodes to a template so values are typed.
• Update existing. Instead of skipping rows whose node already exists, update that node, matched by name or a key column you pick. This makes re-importing a refreshed export safe.

The preview lists how many nodes will be created or updated, plus every skipped row and warning with its row number.

NetBox

NetBox is a popular data-center inventory system. Layertrail reads a NetBox JSON export (the Django dumpdata format). Choose File → Import… and pick the JSON file; Layertrail recognizes it and builds the hierarchy:

Site → Location → Rack → Device

Device, rack, and site attributes become typed properties on generated templates. Devices without a name get NetBox’s own label, <manufacturer> <model> (<id>).

After importing

Save a snapshot right away, for example imported-from-netbox. Every later change can then be compared to the imported baseline.