Table1(
Col1 int not null,
Col2 int not null,
Col3 bigint not null,
Col4 varchar(15) not null,
... other columns...
)
The date is unique on Col1, Col2, Col3 and Col4. and most queries join on those columns. Should I make those the clustered index? Or should I make those non-clustered and I create a clustered index on a new column that is an identity surrogate key?
Thanks,
Dave