Hi,

I think we should not generalize; it is purely on case to case basis, if entity poses ordered attribute and that attribute make sense to the underline domain then we don’t need Identity Column for Cluster Index (assume unique key length is not too large). But at the same time if entity poses time based attribute, then better choice is Identity column (Sedimentary) for Cluster Index.

E.g. – In Banking domain Savings Account Number attribute in Saving entity, here most of the cases data need to be fetch in Sing Account No order. So Saving Account No can use for Cluster index. Here again new account number not creating in sequence, but need to apply some prefix/suffix based on Saving Type or other attribute then we can again go back to Identity column fro Cluster Index.

At the same time Saving Transaction case data normally need to arrange based on time of transaction we can have Identity Column and same can use for Cluster Index.

thanks

Nirmal