SELECT SALARY

FROM    EMPLOYEE

WHERE  SALARY=(SELECT MAX(SALARY)

                         FROM    EMPLOYEE

                         WHERE  SALARY <> (SELECT MAX(SALARY)

                                                      FROM    EMPLOYEE))

The Query is:

SELECT MAX(SAL) AS SAL FROM EMPLOYEE WHERE SAL<>(SELECT MAX(SAL) FROM EMPLOYEE)

SELECT MAX(E1.salary)  FROM ggg E1 , ggg E2
WHERE E1.salary< E2.salary

2.select top 1 salary  from ggg where salary<(select max(salary) from ggg)
order by  salary  desc

3.name   salary
rahul   15000
ronak   20000
saumil  25000

create table ggg (name varchar(20),salary int)

insert into ggg values ('saumil',25000)

select * from ggg

Select MAX(Salary) from ggg
Select MAX(Salary) from ggg where Salary <> (Select MAX(Salary) from ggg)

select top 1 salary  from ggg where salary<(select max(salary) from ggg)
order by  salary  desc

SELECT MAX(E1.salary)  FROM ggg E1 , ggg E2
WHERE E1.salary< E2.salary

doesn't it appear from the volume of code in your solution that it is not optimal. It's clearly like using a sledge hammer to crush a walnut. In the real world no one would bear the overhead of this solution when the others on the page do the same thing without creating tables and certainly without that last join step.

SQL is a non-procedural language, your solution is completely procedural.

Tapas

I don't think that will work because your query is based on the row number and not the value of the salary column. It's the 2nd highest salary, not the row with the 2nd highest rownumber.

In general for a relational database queries involving physical aspects of the table (like rownum) should be avoided.  The relational model is based on comparing values that represent facts about things in the real world, not facts about the internal makeup of tables rows and columns.

This should work

Select Max(Salary)
From tblSalary
Where Salary < (Select Max(Salary) From tblSalary)

Vikram

Hi  u r telling very good answer.

Thanks

Srinivas