From the documentation I see this:
An Incorta Analyzer table is an entity object in a physical schema. It is a derived table built using the Analyzer that queries in-memory data from a physical schema or a business schema. During a load or update job, an Incorta Analyzer table persists to shared storage as Apache Parquet files.
But I'm not sure *when* the parquet file is generated. IA tables don't show up in the load order screen so I'd hope they go last, but if they don't I think they'd need to be built in a separate schema so they could be scheduled to load after the schema(e) referenced by the IA table are updated.
Solved! Go to Solution.
@DustinB helped me out with this one:
"Analyzer tables load after everything: extract, transform, load, joins, formula columns. A side effect of that is you cannot refer to an analyzer table in an MV within the same schema."
I can say from my own experience this holds true. I've had no issues with data load orders with keeping Analyzer tables.