Sunday, January 20, 2008

SQL Optimization

T-SQL Optimization Tips
• Use views and stored procedures instead of heavy-duty queries.
This can reduce network traffic, because your client will send to server only stored procedure or view name (perhaps with some parameters) instead of large heavy-duty queries text. This can be used to facilitate permission management also, because you can restrict user access to table columns they should not see.
• Try to use constraints instead of triggers, whenever possible.
Constraints are much more efficient than triggers and can boost performance. So, you should use constraints instead of triggers, whenever possible.
• Use table variables instead of temporary tables.
Table variables require less locking and logging resources than temporary tables, so table variables should be used whenever possible. The table variables are available in SQL Server 2000 only.
• Try to use UNION ALL statement instead of UNION, whenever possible.
The UNION ALL statement is much faster than UNION, because UNION ALL statement does not look for duplicate rows, and UNION statement does look for duplicate rows, whether or not they exist.
• Try to avoid using the DISTINCT clause, whenever possible.
Because using the DISTINCT clause will result in some performance degradation, you should use this clause only when it is necessary.
• Try to avoid using SQL Server cursors, whenever possible.
SQL Server cursors can result in some performance degradation in comparison with select statements. Try to use correlated sub-query or derived tables, if you need to perform row-by-row operations.
• Try to avoid the HAVING clause, whenever possible.
The HAVING clause is used to restrict the result set returned by the GROUP BY clause. When you use GROUP BY with the HAVING clause, the GROUP BY clause divides the rows into sets of grouped rows and aggregates their values, and then the HAVING clause eliminates undesired aggregated groups. In many cases, you can write your select statement so, that it will contain only WHERE and GROUP BY clauses without HAVING clause. This can improve the performance of your query.
• If you need to return the total table's row count, you can use alternative way instead of SELECT COUNT(*) statement.
Because SELECT COUNT(*) statement make a full table scan to return the total table's row count, it can take very many time for the large table. There is another way to determine the total row count in a table. You can use sysindexes system table, in this case. There is ROWS column in the sysindexes table. This column contains the total row count for each table in your database. So, you can use the following select statement instead of SELECT COUNT(*): SELECT rows FROM sysindexes WHERE id = OBJECT_ID('table_name') AND indid <>
• Include SET NOCOUNT ON statement into your stored procedures to stop the message indicating the number of rows affected by a T-SQL statement.
This can reduce network traffic, because your client will not receive the message indicating the number of rows affected by a T-SQL statement.
• Try to restrict the queries result set by using the WHERE clause.
This can results in good performance benefits, because SQL Server will return to client only particular rows, not all rows from the table(s). This can reduce network traffic and boost the overall performance of the query.
• Use the select statements with TOP keyword or the SET ROWCOUNT statement, if you need to return only the first n rows.
This can improve performance of your queries, because the smaller result set will be returned. This can also reduce the traffic between the server and the clients.
• Try to restrict the queries result set by returning only the particular columns from the table, not all table's columns.
This can results in good performance benefits, because SQL Server will return to client only particular columns, not all table's columns. This can reduce network traffic and boost the overall performance of the query.
1.Indexes
2.avoid more number of triggers on the table
3.unnecessary complicated joins
4.correct use of Group by clause with the select list
5.in worst cases De normalization

Index Optimization tips
• Every index increases the time in takes to perform INSERTS, UPDATES and DELETES, so the number of indexes should not be very much. Try to use maximum 4-5 indexes on one table, not more. If you have read-only table, then the number of indexes may be increased.
• Keep your indexes as narrow as possible. This reduces the size of the index and reduces the number of reads required to read the index.
• Try to create indexes on columns that have integer values rather than character values.
• If you create a composite (multi-column) index, the order of the columns in the key are very important. Try to order the columns in the key as to enhance selectivity, with the most selective columns to the leftmost of the key.
• If you want to join several tables, try to create surrogate integer keys for this purpose and create indexes on their columns.
• Create surrogate integer primary key (identity for example) if your table will not have many insert operations.
• Clustered indexes are more preferable than nonclustered, if you need to select by a range of values or you need to sort results set with GROUP BY or ORDER BY.
• If your application will be performing the same query over and over on the same table, consider creating a covering index on the table.
• You can use the SQL Server Profiler Create Trace Wizard with "Identify Scans of Large Tables" trace to determine which tables in your database may need indexes. This trace will show which tables are being scanned by queries instead of using an index.
• You can use sp_MSforeachtable undocumented stored procedure to rebuild all indexes in your database. Try to schedule it to execute during CPU idle time and slow production periods.
sp_MSforeachtable @command1="print '?' DBCC DBREINDEX ('?')"
QUERIES

1. 2 tables
Employee Phone
empid
empname
salary
mgrid empid
phnumber
2. Select all employees who doesn't have phone?
SELECT empname
FROM Employee
WHERE (empid NOT IN
(SELECT DISTINCT empid
FROM phone))

3. Select the employee names who is having more than one phone numbers.
SELECT empname
FROM employee
WHERE (empid IN
(SELECT empid
FROM phone
GROUP BY empid
HAVING COUNT(empid) > 1))
4. Select the details of 3 max salaried employees from employee table.
SELECT TOP 3 empid, salary
FROM employee
ORDER BY salary DESC
5. Display all managers from the table. (manager id is same as emp id)
SELECT empname
FROM employee
WHERE (empid IN
(SELECT DISTINCT mgrid
FROM employee))
6. Write a Select statement to list the Employee Name, Manager Name under a particular manager?
SELECT e1.empname AS EmpName, e2.empname AS ManagerName
FROM Employee e1 INNER JOIN
Employee e2 ON e1.mgrid = e2.empid
ORDER BY e2.mgrid
7. 2 tables emp and phone.
emp fields are - empid, name
Ph fields are - empid, ph (office, mobile, home). Select all employees who doesn't have any ph nos.
SELECT *
FROM employee LEFT OUTER JOIN
phone ON employee.empid = phone.empid
WHERE (phone.office IS NULL OR phone.office = ' ')
AND (phone.mobile IS NULL OR phone.mobile = ' ')
AND (phone.home IS NULL OR phone.home = ' ')
8. Find employee who is living in more than one city.
Two Tables:
Emp City
Empid
empName
Salary Empid
City
9. SELECT empname, fname, lname
FROM employee
WHERE (empid IN
(SELECT empid
FROM city
GROUP BY empid
HAVING COUNT(empid) > 1))
10. Find all employees who is living in the same city. (table is same as above)
SELECT fname
FROM employee
WHERE (empid IN
(SELECT empid
FROM city a
WHERE city IN
(SELECT city
FROM city b
GROUP BY city
HAVING COUNT(city) > 1)))
11. There is a table named MovieTable with three columns - moviename, person and role. Write a query which gets the movie details where Mr. Amitabh and Mr. Vinod acted and their role is actor.
SELECT DISTINCT m1.moviename
FROM MovieTable m1 INNER JOIN
MovieTable m2 ON m1.moviename = m2.moviename
WHERE (m1.person = 'amitabh' AND m2.person = 'vinod' OR
m2.person = 'amitabh' AND m1.person = 'vinod') AND (m1.role = 'actor') AND (m2.role = 'actor')
ORDER BY m1.moviename
12. There are two employee tables named emp1 and emp2. Both contains same structure (salary details). But Emp2 salary details are incorrect and emp1 salary details are correct. So, write a query which corrects salary details of the table emp2
update a set a.sal=b.sal from emp1 a, emp2 b where a.empid=b.empid
13. Given a Table named “Students” which contains studentid, subjectid and marks. Where there are 10 subjects and 50 students. Write a Query to find out the Maximum marks obtained in each subject.
14. In this same tables now write a SQL Query to get the studentid also to combine with previous results.
15. Three tables – student , course, marks – how do go at finding name of the students who got max marks in the diff courses.
SELECT student.name, course.name AS coursename, marks.sid, marks.mark
FROM marks INNER JOIN
student ON marks.sid = student.sid INNER JOIN
course ON marks.cid = course.cid
WHERE (marks.mark =
(SELECT MAX(Mark)
FROM Marks MaxMark
WHERE MaxMark.cID = Marks.cID))
16. There is a table day_temp which has three columns dayid, day and temperature. How do I write a query to get the difference of temperature among each other for seven days of a week?
SELECT a.dayid, a.dday, a.tempe, a.tempe - b.tempe AS Difference
FROM day_temp a INNER JOIN
day_temp b ON a.dayid = b.dayid + 1
OR
Select a.day, a.degree-b.degree from temperature a, temperature b where a.id=b.id+1
17. There is a table which contains the names like this. a1, a2, a3, a3, a4, a1, a1, a2 and their salaries. Write a query to get grand total salary, and total salaries of individual employees in one query.
SELECT empid, SUM(salary) AS salary
FROM employee
GROUP BY empid WITH ROLLUP
ORDER BY empid
18. How to know how many tables contains empno as a column in a database?
SELECT COUNT(*) AS Counter
FROM syscolumns
WHERE (name = 'empno')
19. Find duplicate rows in a table? OR I have a table with one column which has many records which are not distinct. I need to find the distinct values from that column and number of times it’s repeated.
SELECT sid, mark, COUNT(*) AS Counter
FROM marks
GROUP BY sid, mark
HAVING (COUNT(*) > 1)
20. How to delete the rows which are duplicate (don’t delete both duplicate records).
SET ROWCOUNT 1
DELETE yourtable
FROM yourtable a
WHERE (SELECT COUNT(*) FROM yourtable b WHERE b.name1 = a.name1 AND b.age1 = a.age1) > 1
WHILE @@rowcount > 0
DELETE yourtable
FROM yourtable a
WHERE (SELECT COUNT(*) FROM yourtable b WHERE b.name1 = a.name1 AND b.age1 = a.age1) > 1
SET ROWCOUNT 0
21. How to find 6th highest salary
SELECT TOP 1 salary
FROM (SELECT DISTINCT TOP 6 salary
FROM employee
ORDER BY salary DESC) a
ORDER BY salary
22. Find top salary among two tables
SELECT TOP 1 sal
FROM (SELECT MAX(sal) AS sal
FROM sal1
UNION
SELECT MAX(sal) AS sal
FROM sal2) a
ORDER BY sal DESC
23. Write a query to convert all the letters in a word to upper case
SELECT UPPER('test')
24. Write a query to round up the values of a number. For example even if the user enters 7.1 it should be rounded up to 8.
SELECT CEILING (7.1)
25. Write a SQL Query to find first day of month?
SELECT DATENAME(dw, DATEADD(dd, - DATEPART(dd, GETDATE()) + 1, GETDATE())) AS FirstDay
Datepart Abbreviations
year yy, yyyy
quarter qq, q
month mm, m
dayofyear dy, y
day dd, d
week wk, ww
weekday dw
hour hh
minute mi, n
second ss, s
millisecond ms
26. Table A contains column1 which is primary key and has 2 values (1, 2) and Table B contains column1 which is primary key and has 2 values (2, 3). Write a query which returns the values that are not common for the tables and the query should return one column with 2 records.
SELECT tbla.a
FROM tbla, tblb
WHERE tbla.a <>
(SELECT tblb.a
FROM tbla, tblb
WHERE tbla.a = tblb.a)
UNION
SELECT tblb.a
FROM tbla, tblb
WHERE tblb.a <>
(SELECT tbla.a
FROM tbla, tblb
WHERE tbla.a = tblb.a)

OR (better approach)

SELECT a
FROM tbla
WHERE a NOT IN
(SELECT a
FROM tblb)
UNION ALL
SELECT a
FROM tblb
WHERE a NOT IN
(SELECT a
FROM tbla)
27. There are 3 tables Titles, Authors and Title-Authors (check PUBS db). Write the query to get the author name and the number of books written by that author, the result should start from the author who has written the maximum number of books and end with the author who has written the minimum number of books.
SELECT authors.au_lname, COUNT(*) AS BooksCount
FROM authors INNER JOIN
titleauthor ON authors.au_id = titleauthor.au_id INNER JOIN
titles ON titles.title_id = titleauthor.title_id
GROUP BY authors.au_lname
ORDER BY BooksCount DESC
28.
UPDATE emp_master
SET emp_sal =
CASE
WHEN emp_sal > 0 AND emp_sal <= 20000 THEN (emp_sal * 1.01)
WHEN emp_sal > 20000 THEN (emp_sal * 1.02)
END
29. List all products with total quantity ordered, if quantity ordered is null show it as 0.
SELECT name, CASE WHEN SUM(qty) IS NULL THEN 0 WHEN SUM(qty) > 0 THEN SUM(qty) END AS tot
FROM [order] RIGHT OUTER JOIN
product ON [order].prodid = product.prodid
GROUP BY name
Result:
coke 60
mirinda 0
pepsi 10
30. ANY, SOME, or ALL?
ALL means greater than every value--in other words, greater than the maximum value. For example, >ALL (1, 2, 3) means greater than 3.
ANY means greater than at least one value, that is, greater than the minimum. So >ANY (1, 2, 3) means greater than 1. SOME is an SQL-92 standard equivalent for ANY.
IN & = (difference in correlated sub query)

1 comment:

Anonymous said...

really really nice article