Quick question as to the standards for Data Vault Staging of a Source Table, and should I stage the table broken out by Concept (i.e. Hub).
I have 2 scenarios, the first table, HR is from SQL Server Dispensing Software, and then Contact from Salesforce.
Both have been lost to the art of 3NF.
But, in the HR table, I have Patient Demographics, Doctor, NDC, etc., etc.. Contact has RecordTypes, hence row 1 is a patient, while row 2 is a Doctor. So one break out is by Column (HR) and the other by RecordTypes (Contact), with a few supporting columns to add context to each RecordType.
Is it a best practice to stage the HR broken out into separate staging tables, Patient, Doctor, NDC? I ask, because it's a standard to stage your data with the Hash Key and HashDiff key. If I were to do that as the Source Table is laid out, I could have 6 or more columns just for Hash Keys, plus 6 or more columns just for HashDiff.
Same question would go for the Contact Object in Salesforce.
Thank you all.