Skip to main content

Import templates

An import template sets up an import source for you. Instead of creating types and mapping every stream by hand, you pick the template when you create the source, and Starhive creates a new data space with the types and attributes the source needs, plus a connection for each stream you choose. You can change all of it afterwards.

Templates are available for Lansweeper, Microsoft Intune, and JSM Assets sources. They're rolling out gradually, so they may not be available in every workspace yet.

Set up a source with a template​

  1. Go to Imports, choose Create Source, and pick the source.
  2. Fill in its connection details and choose Create Source.
  3. Starhive asks How should this source be set up? Choose Predefined with new data space. The other option, No predefined configuration, creates nothing: you map each stream to your own spaces and types afterwards, as described in Import (connections).
  4. Starhive checks the template against your source, then lists the streams it can import. Choose which ones you want — All, None, or Default, the template's recommended selection (for JSM Assets, every stream) — or tick them one by one. The summary shows the space, the types to be created, and how many streams will be configured.
  5. Choose Create space and configure source. Creating the types and setting up the connections takes a moment; with many streams, reading your source can take a few minutes.
  6. When it's done, Back To Import takes you to the new source. Run a sync to fill the space.

What the stream list tells you​

Before you apply the template, the stream list points out anything that won't import as you might expect:

  • Not in your source — the template has streams your source doesn't, and they're skipped.
  • Can't be imported — your source lacks the fields a stream's objects are matched on, so the stream can't be imported without creating every object again on each sync. For the Lansweeper asset types and the JSM Assets object types, that stream is simply skipped. For any other stream you've ticked, the template can't be set up until you untick it — the wizard says This cannot be set up for your source and names the stream. The same happens if no stream can be imported at all.
  • References that stay empty — if you leave out a stream that other streams refer to, those references aren't filled. JSM Assets keeps them as text instead, as described below.

What a template creates​

  • A new data space with Restricted permissions. It's named after the kind of source — Lansweeper, Microsoft Intune, or JSM Assets — not the name you gave the source, so rename it if you set up more than one.
  • Types and attributes for what the source holds, with Reference attributes between types where the source's records point at each other.
  • A connection for each stream you chose, with the template's fields mapped and an identifier set, so repeated syncs update the same objects instead of creating new ones. A field your source doesn't have is left unmapped, and so is a reference to a type whose stream you didn't choose (except in JSM Assets, which keeps it as text). Each connection is checked, so it runs on every sync, whether you start it or it runs on a schedule.

After that, the space, its types, and its connections are ordinary ones: rename them, add attributes, change mappings, or switch connections off as you would any other.

Add more streams later​

To import a stream you left out, open the source and choose Create Connection → From Template. Streams the source already imports aren't listed. Each stream you choose gets a type in the template's data space and a connection of its own.

If the source wasn't set up from the template, the streams you choose go into a new data space instead.

If setting it up fails​

If something goes wrong partway, what was created so far is kept, and in most cases trying again carries on from there rather than starting over. The message says what went wrong and whether trying again will help — for example, that your plan doesn't have room for what the template creates, or that your source couldn't be read and its credentials or settings need checking. If it says trying again won't help, for example because the space it was building in has been deleted, contact support.

The templates​

Lansweeper IT assets​

For Lansweeper sources. It creates:

  • A type for each kind of asset your Lansweeper workspace has — Windows, Linux, Laptop, Switch, Printer, and so on — each extending a shared Asset type that holds the hardware and ownership details every asset has.
  • Operating system and Software, with each software title referencing the operating systems it runs on.
  • Vulnerability, with each vulnerability referencing the assets it affects.

The default selection covers the common asset types, operating systems, and software; vulnerabilities are left for you to tick. Only Lansweeper's own documented fields are mapped — custom fields you've defined in your Lansweeper tenant aren't.

Intune device inventory​

For Microsoft Intune sources. It creates managed devices, each referencing its primary user, category, manufacturer, model, operating system, and operating system version, and the apps Intune has detected, each referencing the devices it was found on.

Windows Autopilot devices, published apps, app configurations, compliance policies, and device configurations can be ticked as well.

The app registration needs the permissions listed on the Intune import page. Users need User.Read.All: without it, no users are imported and devices don't reference their primary user. Autopilot also needs DeviceManagementServiceConfig.Read.All, which the other streams don't, so it's only imported if you tick it.

A device that has never contacted Intune shows its last contact as 1 January 0001.

JSM Assets objects​

For JSM Assets sources. It creates a type for every object type in your Assets object schemas, with the attributes its owner gave it:

  • Your schemas or one of them. By default, every object schema the API token can read goes into one JSM Assets space, with each type named after its schema and object type ("IT - Laptop"). Set Object Schema ID on the source to import a single schema instead: the space is named after the schema, and the types after the object types alone.
  • The same tree as the Assets menu. Each type sits under its parent object type. A type that inherits from its parent in Assets extends its parent's type, so an inherited attribute exists once, on the type that defines it. An abstract object type becomes an abstract type, with nothing of its own to import.
  • References stay references. Each Assets reference attribute becomes a Reference attribute pointing at the type made from the referenced object type. It keeps the referenced objects' keys as text instead when that object type isn't imported — its stream left out or unable to be imported — and where references would go round in a circle: a type pointing at itself, such as a parent location, or two types pointing at each other. A circle would make its types sync last, with a warning on every run, so one of its references stays text. An attribute that was set up as text stays text, even if you add the referenced object type's stream later.
  • Other attributes are text. Only you know what your attributes hold, and a value a type refuses would fail the whole object. The attribute an object type uses as its label isn't repeated, since it's the object's Name, and an attribute with a dot in its name isn't imported.

Every object also gets its Object ID, Object key, Name, and when it was created and last updated in Assets. Objects are matched on their Assets object ID rather than their key, so changing a schema's key prefix doesn't import its objects again.