Question

I had asked earlier in SO, how to wrap the multiple column in single column, regarding the same, but i picked up another approach and done that part. Now the requirement has been slightly changed and become, if any of the field in table is update then the query or trigger will be able to detect – on what table has and what columns.

Update

For example I have table tblAddress having column Address, WorkPhone, CellPhone, City, State etc. If a user updates WorkPhone then the query should be able to detect that in table address phone number has been updated and the message will be inserted to another table called tblHistory where tblHistory is mine audit-trail maintaining table and tblAddress is where insert update or delete operation will be performed.

Answer 1

Well, there are two basic patterns you can use:

1. Store all of the column values in a history table (e.g. tblAddress_History) for each audited table (e.g. tblAddress) when a row in the audited table is updated. Each history table would basically mirror the schema of its corresponding audited table (plus maybe a comment or audit info such as when and by whom the update was made).
2. Store a separate row for individual columns in a history table (e.g. tblHistory) for all audited tables; when a row is updated you’d need to create a separate row in the history table for each column that is updated. Your history table would then need to include columns to identify the audited table and the updated column.