Asset Ingestion, Processing, and Metadata
Ingestion has several entry points: the Assets UI upload, desktop app, API (HTTP//api/assets), and bulk import. Each triggers processing. In on-prem/AMS, the DAM Update Asset workflow orchestrates: extract metadata (via XMP extraction), generate thumbnails/renditions, run any configured post-processing. In Cloud Service, the Asset Compute Service and microservices handle extraction and rendition generation asynchronously.
Metadata is the highest-yield exam topic. Know the difference between embedded XMP (inside the file), sidecar XMP (a .xmp file alongside the asset), and repository metadata (the metadata node). The XMP write-back workflow syncs repository edits back into the binary. Metadata schemas define which fields appear in the editor; they are configured under /conf/global/settings/dam/adminui-embedded/metadata (or via the Metadata Schemas console).
Processing profiles are the modern mechanism to control renditions and metadata at upload time - they replace ad-hoc workflow launchers and are the correct answer for "how do I apply a custom rendition to everything uploaded to folder X." Traps: (1) assuming a processing profile edits embedded XMP automatically - it applies the profile's metadata mapping, not arbitrary XMP; (2) confusing metadata schema (UI fields) with metadata mapping (data transformation); (3) forgetting that extraction is asynchronous, so a just-uploaded asset may briefly lack metadata.
Worked rule: to add a custom watermark to all images in a folder - create a processing profile with a custom rendition step referencing an Asset Compute worker (cloud) or a workflow (on-prem), apply it to the folder, then re-process existing assets.