Question

I have a table called Transactions with the following schema:

ColumnName     DataType             Constraints
----------    ----------            -----------
   id             int           PK (Cluster-Index)
 details      varchar(50)

Later on, I add the following two columns to this table:

ColumnName     DataType             Constraints
----------    ----------            -----------
   id             int           PK (Cluster-Index)
 details      varchar(50)
  date          datetime 
 comment      varchar(255)

What will be the index performance on that table with the following query?

select * from transactions where id=somenumber

Will the performance change because I added the two columns? Will there be an effect on the clustered index?

Answer 1

Your performance will roughly be the same. Your clustered index defines the physical ordering of the rows. When you do a query on a clustered primary key, the database essentially does a binary search for your data. The result of adding columns means that not as many rows fit on the same data page. This means the database may have to do a bit more IO to get the same data.