|
Hi all,
how to use query optimization in sql server 2005 for an efficient query. 
Ash
modified on Friday, March 28, 2008 1:48 AM
|
|
|
|
|
Hi Ash,
Here are some of the methods that most of the DBA's use to optimise the speed of MS SWL Server queries:
-Display the minimum number of fields in a query. Set criteria dependant fields that are not required in the dynaset to "not shown".
-Index all restriction based fields, all fields included in expressions, all sorted fields and all join fields.
-Use primary keys or unique indexes wherever possible.
-Use numeric rather than text primary keys.
-Use non blank unique fields.
-Avoid the use of IIf() function in queries.
-Avoid domain aggregate functions such as Dlookup().
-Make careful use of Between and Equal to, rather than > or < speeds up queries.
-Use fixed column headings in Crosstab queries.
-For reports based on queries use Portrait view in preference to Landscape and select Fast Laser
Printing to Yes (View,Options,Other Properties).
-Use Make table queries for running reports on static data. These are called snapshot reports.
-Use Count (*) rather than Count(Column).
-When creating restrictions on a joined column in one-to-many relationships, test out the comparative performance when placing the restriction on the "one" side or the "many" side. The "one" side is not always the fastest - the "many" may have markedly fewer records.
-Short table and field names run faster than long names.
-Normalise tables - join strategies execute more quickly on smaller tables.
-Denormalise tables - reduce the number of joins. Get the balance right between normalisation and denormalisation by experiment.
-Avoid the use of Distinct Row queries - Union queries do not need the distinct row feature as they are automatically return unique fields unless set to Union All.
You may visit the following link for more details about queyr optimization techniques.
http://blogs.msdn.com/QueryOptTeam/[^]
Hope this helps .
Regards,
John Adams
ComponentOne LLC
|
|
|
|
|
That was a good answer!! covered lots of things.
thanks for sharing
Believe Yourself™
|
|
|
|
|
Hi thanks for looking
I have a weird situation here
im doing a case on a column, i need to get the final price
Price = 15.0000
Discount = 10
when '%' then dbo.vwProducts.Price * ((100 - dbo.Promos.Discount) / 100)
Shouldnt this give a result of 13.5?? (15 * ((100-10)/100))
It gives .0000
If i change the 100 for a 10 it works!! It gives a result of 135.0000
(15 * ((100-10)/10)) = 135.0000
Whats happening??
Alexei Rodriguez
|
|
|
|
|
Check your data types. I just tried it with price and discoutn as decimal and it worked. I changed discount to int and got zero.
Hope this helps
Bob
Ashfield Consultants Ltd
|
|
|
|
|
Hi,
You are doing division and multiplication through only int datatype only.
Do type-casting of one of the variables or value.
say (price * ((100-10) / 10.0))
Hope this shud work 4 U...
Regards
SG
|
|
|
|
|
Datatypes were were the problem
Thanks a lot
Alexei Rodriguez
|
|
|
|
|
I am looping through a recordset, and I am both creating and destroying a QueryDef with what I believe to be a pass-through query. The records get processed correctly most of the time but then I get this abberant Error #3218 which states:
"DAO.QueryDefs. Could not update. Currently locked."
What does this mean and how do I prevent it from occurring? The application works correctly most of the time, but this error occurs intermittently, thus, it is the hardest cause to identify. Why is it locked up, and what is locked up?
|
|
|
|
|
Error 3218 "Could not update; currently locked."
This error occurs when a user tries to save a record that is locked by another user.
To handle this error, program your solution to wait for a short period of time, and then try to save the record again. Or, you can display a message that explains the problem and give users the opportunity to try the operation again.
|
|
|
|
|
I am trying tie a reporting services data source with a custom report item and can't seem to find any helpful samples/documentation anywhere. This task is taking much longer than it should have at this point and I could really use some help if there's anyone who has done this before.
What I want to do is create a series designer properties which allow me to :
1) select a data source from a list of available sources which have been defined for the report
2) select fields from the data source and assign them to designer properties
3) navigate the CustomReportItem.CustomData (is that the correct object?) object using the field names which were selected in step 2 (above) to retrieve the runtime values and bind them to my custom report item.
|
|
|
|
|
A) In the CustomReportItemDesigner class (your custom class which inherits from it as a base), override the InitializeNewComponent() method like this:
<br />
public override void InitializeNewComponent()<br />
{<br />
base.InitializeNewComponent();<br />
<br />
CustomData = new CustomData();<br />
<br />
<br />
CustomData.DataRowGroupings = new DataRowGroupings();<br />
CustomData.DataRowGroupings.DataGroupings = new List<DataGrouping>(1);<br />
CustomData.DataRowGroupings.DataGroupings.Add(new DataGrouping());<br />
CustomData.DataRowGroupings.DataGroupings[0].Static = true;<br />
<br />
CustomData.DataColumnGroupings = new DataColumnGroupings();<br />
CustomData.DataColumnGroupings.DataGroupings = new List<DataGrouping>(1);<br />
CustomData.DataColumnGroupings.DataGroupings.Add(new DataGrouping());<br />
CustomData.DataColumnGroupings.DataGroupings[0].Static = true;<br />
<br />
CustomData.DataRows = new List<DataRow>(1);<br />
CustomData.DataRows.Add(new DataRow());<br />
CustomData.DataRows[0].Add(new DataCell());<br />
<br />
CustomData.DataRows[0][0].Add(new DataValue("Label", String.Empty));<br />
CustomData.DataRows[0][0].Add(new DataValue("X", String.Empty));<br />
CustomData.DataRows[0][0].Add(new DataValue("Y", String.Empty));<br />
}<br />
B) Define custom designer properties to set the expressions for the DataValuesCollection (only showing one property here to keep things shorter)
<br />
[Browsable(true), Category("Data")]<br />
public string XValue<br />
{<br />
get<br />
{<br />
if (CustomData.DataRows.Count > 0 && CustomData.DataRows[0].Count > 0)<br />
return GetDataValue(CustomData.DataRows[0][0], "X");<br />
else<br />
return "X Coordinate";<br />
}<br />
set<br />
{<br />
SetDataValue(CustomData.DataRows[0][0], "X", value);<br />
}<br />
}<br />
<br />
<br />
private string GetDataValue(DataCell cell, string name)<br />
{<br />
foreach (DataValue value in cell)<br />
if (value.Name == name)<br />
return value.Value;<br />
return null;<br />
}<br />
<br />
private void SetDataValue(DataCell cell, string name, string expression)<br />
{<br />
foreach (DataValue value in cell)<br />
if (value.Name == name)<br />
{<br />
value.Value = expression;<br />
return;<br />
}<br />
DataValue datavalue = new DataValue(name, expression);<br />
cell.Add(datavalue);<br />
}<br />
C) Make sure your designer provides a property to set the data set name (CustomData.DataSetName).
D) Your CustomReportItem class can then read the data like this:
<br />
private void ReadCustomData()<br />
{ <br />
DataCellCollection rows = m_CRI.CustomData.DataCells;<br />
<br />
int xpos = -1, ypos = -1, labelpos = -1;<br />
<br />
for (int field = 0; field < rows[0, 0].DataValues.Count; field++)<br />
{<br />
switch (rows[0, 0].DataValues[field].Name)<br />
{<br />
case "X":<br />
xpos = field;<br />
break;<br />
case "Y":<br />
ypos = field;<br />
break;<br />
case "Label":<br />
labelpos = field;<br />
break;<br />
}<br />
}<br />
<br />
for (int row = 0; row < rows.RowCount; row++)<br />
{<br />
DataValueCollection values = rows[row, 0].DataValues;<br />
<br />
double x = Convert.ToDouble(values[xpos].Value);<br />
double y = Convert.ToDouble(values[ypos].Value);<br />
string label = (string)values[labelpos].Value;<br />
}<br />
}<br />
NOTE: This assumes the data in your data set is already grouped and the report does not need to perform any grouping on the data in your dataset.
|
|
|
|
|
I'm Working with Asp.Net with C#.. & I'm Generating Reports in SSRS-2005..
Till Now I'm Generating Reports in SSRS-2005 with Stored Procedure.. in Which I'm Generating Reports for One Particular User Details(ie I wrote Stored Procedure for Getting One Users Details)..
Now I Want to Generate the Reports for Different Users DYNAMICALLY (Here, I wrote another Stored Procedure which contains UsersID's as Input Parameters.. when we give the User ID then that Particular User's Details will be Executed).. Here I want to Place one Dropdownlist Control which contains All the User's ID's.. After selecting the Particular User ID, I Want to Generate that User's Details REPORT..
I done this as like that I generated Report for One User(ie as like First Report).. But it is Not Working..
Please give me the Suggestions for this...
Thank You..
|
|
|
|
|
A good while ago there was a lot of fuss made about not using this prefix. Is this still relevant?
|
|
|
|
|
Hi,
Our head SQL guy insists on using the 'sp' as a prefix ... or at the very least maybe two characters prefixing the 'sp' prefix related to the project.
it just makes sense.
Jammer
Going where everyone here has gone before!
My Blog
|
|
|
|
|
My motivation is somewhat unusual: I'm prefixing all application procedures with ap and all system procedures with sp, where application procedures access data that changes during normal tasks, and system procedures access data that only changes during system configuration.
|
|
|
|
|
Jammer wrote: it just makes sense.
Does it? If so, why?
Isn't Hungarian notation[^] a bit passe these days? In Management Studion, stored procedures are listed under their own section so you don't need a naming convention to distinguish them from other objects. Also, stored procedures have a different object type in the SQL Server metadata so you can find all instances of them without having to search on a particular prefix.
Naming stored procedures based on what they do, not what they are seems the most sensible approach.
Paul Marfleet
"No, his mind is not for rent
To any God or government"
Tom Sawyer - Rush
|
|
|
|
|
Agreed. I find hungarian notation just unnecessary ugly keystrokes. And I don't like putting all SQL statements in upper case, but I seem to be alone in my dislike of that.
Regards,
Rob Philpott.
|
|
|
|
|
This limitation has nothing to do with Hungarian Notation, and everything to do with maximising SQL Server performance. See my post below.
|
|
|
|
|
This article relates to a problem caused by prefixing stored procedure names with 'sp_'. The OP was talking about the prefix 'sp', which presumably doesn't cause the same problem.
My issue is with the practice of using unnecessary naming conventions. Prefixing a stored procedure with 'sp' adds no value. You can tell by the context in which it is used that a database object is a stored procedure. What is useful is to choose a name which indicates the its basic purpose.
Paul Marfleet
"No, his mind is not for rent
To any God or government"
Tom Sawyer - Rush
|
|
|
|
|
As I said above, yy prefix is not to indicate that it is a stored procedure, but to indicate a system procedure, one which access data not changed by the business processes of the application, such as lists for drop down etc. compared to 'transactional' data inserted by application procedures, prefixed with ap. My initial idea was these prefixes themselves prefixed with an acronym for the application name, but hosting objects for different applications in the same database is not one of my good practices. Databases are relatively cheap, and and cohesive application should have its own database.
|
|
|
|
|
Actually, the OP was asking about the fuss over the sp prefix, which was the fuss over using sp_. It may seem like splitting hairs, but there was never a problem with just using sp (and I don't think anybody ever raised a fuss over it).
pmarfleet wrote: My issue is with the practice of using unnecessary naming conventions
Agreed. Mindless mantra is annoying.
|
|
|
|
|
I think what you may be referring to is that if you prefix a stored procedure with 'sp_' SQL Server will look up all the extended stored procedures first to see if there is a match.
Apparently on large OLAP/OLTP systems this can result in a significant use of processor power.
I have named my stored procedures 'sp_'and am in the process of changing them and encouraging others not to use them.
I think at the end of the day unless your database is having to process tens to hundreds of millions of records via a stored procedure you should be ok with using 'sp_' as the prefix.
This is just an opinion and as you may discover DBA's tend to be very opinionated (except they see themselves as being 'right'  ).
You always pass failure on the way to success.
|
|
|
|
|
Yup. There is a performance penalty when using them in a database other than Master. There's a technical reason that this happens, which you can find here[^]. As far as I can tell, even in SQL 2008, this is still the case.
|
|
|
|
|
Excellent. I like that, let there be no dispute that this is bad pratice.
Regards,
Rob Philpott.
|
|
|
|
|
call your procs whatever you like, but I think it's dumb. Of course is a sp. SQL Server management studio does that grouping for you.
Christian Graus
Please read this if you don't understand the answer I've given you
"also I don't think "TranslateOneToTwoBillion OneHundredAndFortySevenMillion FourHundredAndEightyThreeThousand SixHundredAndFortySeven()" is a very good choice for a function name" - SpacixOne ( offering help to someone who really needed it ) ( spaces added for the benefit of people running at < 1280x1024 )
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
