I’m doing some work with nested json pipelines into a procedure. I see that the new release is capable of faster json analytics. My question is: does this apply to data ingestion? I bring the data into table/column format rather than a json column.
We improved by a lot the performance of loading sparse JSON data in 8.1. That means JSON with a tremendous number of distinct properties spread across all the documents in a column, but not that many distinct properties per document.
We speeded up JSON seeking by up to 400x in 8.0.
Load speed for non-sparse JSON documents should be comparable to 8.0.