Exam 70-767 topic 1 question 126 discussion

How can be more than one clustered index in the table?

Prevent duplicate rows from being entered by a single system/Allow for point lookups of a single sale StoreId+SaleId: This combo would meet the above requireptments. Each sale generates a receipt, that by law must identity unqiuley a sale transaction. A store should not have duplicate receipt id's. The column SystemID is a red hering(personal point of view):" The company has many store locations. Each location runs their own independent sales systems." This can mean that two separate locations use the same "independent sales system", thus a combo of StoreID+SystemId does not guranatee uniquness. Safest combo would be StoreId+SaleID+SystemID, but when building a compoiste key we should stick to as few columns as necesary. Store rows in the order that they are inserted into the table: saleid: clustered index on it. Because it is an identity(1,1) column, oredring the table by it enforces the above request

SalesID : primary key column SystemID : clustered key column ReceiptID : clustered key column StoredID : primary key column