Returns errors instead of just logging them in the Create/Update match list items methods, so that when something like a BAD_REQUEST occurs, we do not timeout while waiting for a successful state change confirmation that never happens.
Before, updating match list items worked by deleting all old items and then adding all new items. This caused a bug where the terraform plan/apply, based on the item ids computed/returned by the API, was somehow determining that new items were updates to old items. It then POSTed a BAD_REQUEST with invalid id params by trying to provide the old ids. This fixes it so that we instead:
DELETE - delete old items that aren't being updated
PUT - update old items with new item values
POST - add remaining new items that aren't updates
INVS-2114
id
params by trying to provide the old ids. This fixes it so that we instead: