What are the advantage/ disadvantages of using a primary key that is a combination of multiple columns?
I have come across a table that has a unique non clustered index as a GUID but the clustering index is applied on the combination of 9 columns. In total the table has 29 columns.
Can anyone help me understand what might be the benefits of having such a clustering key? As per Kimberly Tripp || Index Debate by Kimberly Tripp the clustering Index should be narrow, unique and static, but this case is totally opposite to what Kimberly states:
- the primary key is susceptible to changes so its not static
- It is definitely not narrow as it spans multiple columns and other non clustering indexes refer to it for mapping. so it is definitely expensive for IAM/PFS.