Question

I am new to SQL and trying to achieve something.

I have a table having fields like:

CustId   Fname   Lname    Address1    Address2  Zip    State  LoadDate   Flag
1        John     Leo      xxx        xxx       34532    VA    1/1/2012   Y
2        Mike     Saunder  xxx        xxx       94090    CA    1/1/2012   Y

Everytime I insert a new record it will set its flag to Y but lets say if any of the record value changes then new one will be set to Flag Y and old one to N. So for example if Mikes address1 is changed.

CustId   Fname   Lname    Address1    Address2  Zip    State  LoadDate   Flag
1        John     Leo      xxx        xxx       34532    VA    1/1/2012   Y
2        Mike     Saunder  xxx        xxx       94090    CA    1/1/2012   N
3        Mike     Saunder  rrr        xxx       94090    CA    2/1/2012   Y

Answer 1

Assuming that you’re trying to say that the Fname/Lname pair is a unique key that lets you match rows:

declare @Folks as table ( CustId int, Fname varchar(10), Lname varchar(10),
  Address1 varchar(10), Address2 varchar(10),
  Zip varchar(9), State varchar(2), LoadDate date, Flag char(1) )
insert into @Folks ( CustId, Fname, Lname, Address1, Address2, Zip, State, LoadDate, Flag ) values
  ( 1, 'John', 'Leo', 'xxx', 'xxx', '34532', 'VA', '1/1/2012', '-' ),
  ( 2, 'Mike', 'Saunder', 'xxx', 'xxx', '94090', 'CA', '1/1/2012', '-' ),
  ( 3, 'Mike', 'Saunder', 'rrr', 'xxx', '94090', 'CA', '2/1/2012', '-' )

-- Before setting the Flag.
select * from @Folks

-- Computing a flying flag rather than using a stored value.
select *, case when LoadDate = ( select MAX(LoadDate) from @Folks as S where S.Fname = O.Fname and S.Lname = O.Lname ) then 'Y' else 'N' end as 'FlyingFlag'
  from @Folks as O

-- Updating all stored values.  Every bloody row.
update @Folks
  set Flag = case when LoadDate = ( select MAX(LoadDate) from @Folks as S where S.Fname = O.Fname and S.Lname = O.Lname ) then 'Y' else 'N' end
  from @Folks as O

-- Showing the result.
select * from @Folks

EDIT:
I still think this is not the direction you want to be heading. For example, it doesn’t gracefully accomodate more than one update per customer per day unless you rely on the CustId to be a tie breaker.

A marginally better solution would be to replace Flag with a PreviousId field that is NULL in the first row for a customer and in later rows contains the CustId of the row that is being superseded. That links together the changes for each customer and provides a well defined order in the event that a customer is updated repeatedly on a single date.