We are using the bulk data api and in thread about fields,https://community.lithium.com/t5/Developers-Discussion/Bulk-Data-API-Too-Many-Fields-Error/m-p/21811..., the PM recommended to use document ID as the key to match up data with updates. Our BI guy is wondering if others are using the document ID with success to merge data with bulk data updates (adds columns, backfills data). From what I understand the document.id as it not as efficient as a unique number or unique date. I asked about using the create date, but the BI concern is the unix datetime has the potential for the same date, and instead would like a unique date on database insert or a number. Are other bulk data api users successfully using document.id across millions of recording when updating that data with inserts?