I want to know the # of users on our web site for each month in a given year. I'm looking for a faster way to do this--perhaps one that can leverage an index instead of reading the entire table! (My avg disk queue right now is above 7 and the query takes about 90 seconds).
Here's my current SP. Basically I'm calculating each month/year and using UNION to join them together, then pivot to rotate.
hi all, if i have a comma delimited string and want to insert each delimited substring into a table which of the following way is faster?pass the whole string into the a stored procedure and loop through the delimited string and pick out the substring and insert into the table orloop and pass the substring into a stored procedure and insert N times?or any other better ways someone could suggest me to do thanks!
I was just wondering if this can be done any faster? code-wise that is...
Don't mind the converts, can't do without them, as the data discipline for the source table isn't always reliable, while I have to be absolutely sure the destination data ends in the required format.
I have SQL file but it run slowly when comes to huge record. How do I make it faster. I do create an index but how to make use the index? Pls help me on this...
I'm sonewhat new to MS SQL Server and I'm wondering about which of thefollowing two queries would be faster:DECLARE @ResidencesBuilt intDECLARE @BarracksBuilt intDECLARE @AirBaysBuilt intDECLARE @NuclearPlantsBuilt intDECLARE @FusionPlantsBuilt intDECLARE @StarMinesBuilt intDECLARE @TrainingCampsBuilt intDECLARE @FactoriesBuilt intSELECT@ResidencesBuilt = SUM(CASE WHEN BuildingType = 0 THEN Built END),@BarracksBuilt = SUM(CASE WHEN BuildingType = 1 THEN Built END),@AirBaysBuilt = SUM(CASE WHEN BuildingType = 2 THEN Built END),@NuclearPlantsBuilt = SUM(CASE WHEN BuildingType = 3 THEN Built END),@FusionPlantsBuilt = SUM(CASE WHEN BuildingType = 4 THEN Built END),@StarMinesBuilt = SUM(CASE WHEN BuildingType = 5 THEN Built END),@TrainingCampsBuilt = SUM(CASE WHEN BuildingType = 6 THEN Built END),@FactoriesBuilt = SUM(CASE WHEN BuildingType = 7 THEN Built END)FROM BuildingsGROUP BY kdIDHAVING kdID = 2902Or:DECLARE @ResidencesBuilt intDECLARE @BarracksBuilt intDECLARE @AirBaysBuilt intDECLARE @NuclearPlantsBuilt intDECLARE @FusionPlantsBuilt intDECLARE @StarMinesBuilt intDECLARE @TrainingCampsBuilt intDECLARE @FactoriesBuilt intSET @ResidencesBuilt = (SELECT Built FROM Buildings WHERE BuildingType = 0AND kdID = 2902)SET @BarracksBuilt = (SELECT Built FROM Buildings WHERE BuildingType = 1 ANDkdID = 2902)SET @AirBaysBuilt = (SELECT Built FROM Buildings WHERE BuildingType = 2 ANDkdID = 2902)SET @NuclearPlantsBuilt = (SELECT Built FROM Buildings WHERE BuildingType =3 AND kdID = 2902)SET @FusionPlantsBuilt = (SELECT Built FROM Buildings WHERE BuildingType = 4AND kdID = 2902)SET @StarMinesBuilt = (SELECT Built FROM Buildings WHERE BuildingType = 5AND kdID = 2902)SET @TrainingCampsBuilt = (SELECT Built FROM Buildings WHERE BuildingType =6 AND kdID = 2902)SET @FactoriesBuilt = (SELECT Built FROM Buildings WHERE BuildingType = 7AND kdID = 2902)The data source is:kdID BuildingType Built2902 6 02902 7 02902 4 02902 0 802902 2 02902 1 52902 3 402902 5 10Or:CREATE TABLE [dbo].[Buildings] ([kdID] [int],[BuildingType] [tinyint],[Built] [int])INSERT INTO Buildings (kdID, BuildingType, Built) VALUES (2902, 0, 80)INSERT INTO Buildings (kdID, BuildingType, Built) VALUES (2902, 1, 5)INSERT INTO Buildings (kdID, BuildingType, Built) VALUES (2902, 2, 0)INSERT INTO Buildings (kdID, BuildingType, Built) VALUES (2902, 3, 40)INSERT INTO Buildings (kdID, BuildingType, Built) VALUES (2902, 4, 0)INSERT INTO Buildings (kdID, BuildingType, Built) VALUES (2902, 5, 10)INSERT INTO Buildings (kdID, BuildingType, Built) VALUES (2902, 6, 0)INSERT INTO Buildings (kdID, BuildingType, Built) VALUES (2902, 7, 0)Analyzer says the first would be faster, but it has a lot of SUM()'s andwhatnot so I'm not too sure about this. There are also about 1000 rows inthe actual Buildings table. This will be a part of a stored procedure.
Hi all, I m new to this forum and this is my first question. I m having 2 pages in my web site ... page 1 query directly to db using sqldatasource, the second page query through a BLL then DAL by following the step in this tutorial (http://www.asp.net/learn/dataaccess/tutorial02vb.aspx?tabid=63).... Page 1 is using a "Like" query search and the Page 2 is the normal displaying some product detail.... Under normal circumstances, one will expect Page 1 will be way fastest than the Page 2... however the problem is Page 1 is in thunder speed while Page 2 takes 10 secs to load... 10 seconds is really not acceptable... I really couldnt figure out what happens... both Page 1 and Page 2 are using the same connection string which connection through a DSN.... How is the connection different by using sqldatasource and DAL?? Could someone please help.... ThanksP.S. I m using a Pervasive database
Hi y'all, I've recently run a profiler on my code and following query took 7 seconds: SELECT TOP 10 UI, COUNT(UI) AS Expr1 FROM table WHERE (UI <> 'custom_welcome') GROUP BY UI ORDER BY COUNT(UI) DESC Is it possible to rewrite this so my code gets faster? It's also possible that it's due to the size of the table? Thanks in advance! I'll let you know how long your query takes :)
I have a cursor prcedure that is pretty slow because as the cursor moves through the data I have three select statement on the same table to find other rows information. Is there a better way to do this? Simple Example of Code is: DECLARE MyVARABLES 1 to X DECLARE c1 CURSOR FOR SELECT MyData1, MyData2 to X FROM MyTable FOR UPDATE OF MyUpdateData --Start Cursor OPEN c1 FETCH NEXT FROM c1 INTO MyVariables --LOOP WHILE @@FETCH_STATUS = 0 BEGIN ----------------------- -- Get other rows data to add to this rows data ......GUESSING THIS IS THE SLOW PART as the table is LARGESELECT MyVar1 = MyData1 FROM MyTable WHERE MyTableColumns = MyVariables AND MyTableColumns2 <> MyVariables2 --FINDS OTHER ROW (I have three of these)
--Calculate & Update If MyVarable = 'this or that' BEGIN UPDATE MyTable SET MyUpdateData = MyVar1 * x *y WHERE CURRENT OF c1 END ------------------- -- NextFETCH NEXT FROM c1 INTO MyVarables 1 to xEND CLOSE c1 DEALLOCATE c1
Hi! I M basically an application developer & use simple sql queries in my programmings. I do not have much idea abt tuning/auditing part & thatswhy i m unable to answer them properly in my interviews. Can anybody give me some tips?????
Question 1: In a stored procedure, One SELECT stmt is there & depending upon the @rowcount, it updates around 14000 records which is also written inside this stored procedure. Instead of writing this way, there is some other way which is faster than this. Can anybody tell me the correct way???
Question 2:Can anybody give me few examples like this?????? I need them desparetly.
I´ve created a class to make some standard transaction development a little bit faster. The destructor seem to run, but something makes this object slow down the database, if SqlTransaction and/or SqlConnection isnt manualy handled with the method Commit(). Any ideas on how to handle the SqlTransaction and SqlConnection better?
public class DataTransaction { private bool blnError = false; private ArrayList arrErrorList = new ArrayList(); private SqlConnectionobjConnection = new SqlConnection(System.Configuration.ConfigurationSettings.AppSettings["ConnectionString"].ToString()); private SqlTransactionobjTransaction;
Previosuly I was executing 2 DTS packages one afte the other manually and together they took a CONSIDERABLE time. The 1st one was pulling data from the OLPT, doing transformations and populating the tables in my Datamart and the 2nd one was doing a FULL process of all the dimensions and cubes.
However I tried scheduling the DTSs as jobs and havethen merged the 2 resulting jobs as a SINGLE job having 2 sequential steps. To my surprise the resulting job takes less than half the time (actually even lesser) as compared with my original approach i.e. running the DTSs. And I am talking about major improvement in terms of completion of the tasks here :)
Am i getting over excited here or is this natural? I assume that if this is correct then jobs much be some sort of "compiled" version as compared to DTS and maybe that's why I have this terrific improvement in terms of execution times.
I have rewritten a stored procedure that consists of a single select that selects from a view. Essentially I combined the select in the view and the select in the sp into one select. I am now trying to determine if the new version is faster.
The estimated execution plan gives a ratio of 96% : 4% in favour of the new version when I run them together from a query window but when I try to time them I can't get a satisfactory result.
If I run each query once and display the difference between start and end time, they display 0. If I run each one 100, 200, etc times I get different results each time.
If I had a WHERE clause that had to compare a string to another string would it be faster one way or another if I broke it down to three different, smaller searches?
Previosuly I was executing 2 DTS packages one afte the other manually and together they took a CONSIDERABLE time. The 1st one was pulling data from the OLPT, doing transformations and populating the tables in my Datamart and the 2nd one was doing a FULL process of all the dimensions and cubes.
However I tried scheduling the DTSs as jobs and havethen merged the 2 resulting jobs as a SINGLe job having 2 sequential steps. To my surprise the resulting job takes less than half the time (actually even lesser) as compared with my original approach i.e. running the DTSs.
Am i getting over excited here or is this natural? I assume that if this is correct then jobs much be some sort of "compiled" version as compared to DTS and maybe that's why I have this terrific improvement in terms of execution times.
I have a live database and an archive database. I update the archive tables once a day from the live tables using:
INSERT INTO arc_table SELECT * FROM cur_table AS cur WHERE NOT EXISTS (SELECT * FROM arc_table AS arc WHERE arc.key = cur.key)
GO
This inserts newer records into the archive tables from the live tables.
I have two different methods to clean the live tables once a week but keep data from the previous week. Both methods have been verified to delete the same rows.
DELETE cur_table WHERE EXISTS (SELECT key FROM arc_table AS arc WHERE arc.key = cur_table.key) AND date_time < GetDate() - 7
GO
Second method modified from BOL - deletes identical rows
DELETE cur_table FROM (SELECT key FROM arc_table) AS arc WHERE arc.key = cur_table.key AND date_time < GetDate() - 7
GO
I read that "WHERE [NOT] EXISTS" is faster than "WHERE [NOT] IN" but this is the first time I have seen DELETE xx FROM (SELECT ----)
I'd like to know which procedure will be faster and/or better.
HelloI need this really faster in mS SQL 2000Usernumber (int)reportid (FK)reportreportid (PK)Category (int)SELECT A, B, C, D INTO UserCopy FROM UserWHERE User.reportid IN (SELECT MAX(report.reportID) AS maxReport FROM Report GROUP BY report.Category) AND user.number NOT IN (120,144,206,345,221,789,548,666,1204,4875,22,135, 777,444)can return a more than 1000 rows (an the table = 10.000 rows): SELECT MAX(report.reportID) AS maxReport FROM Report GROUP BY report.Categoryand the table user has a few millions rowsReport.ReportId is a Primary key for User.reportid (FK) for the moment it takes up to 3 minutes, i need to do that in 30 seconds maximumthank you for helping
I recently installed SQL2005 Express on a Dell Precision workstation to be accessed by 6 users. The PC has XP Pro, 2GB memory, and 2.0GHz core2due CPU. This PC is dedicated for the database only. I also set XP Performance options in System Properties to Background services and System cache. Now, the users are experiencing slow response from the database. How can I make the database response faster? Thanks for your input.
Can anyone give me the basics of speeding up reports that use queries or views or nested views? Current reports are now taking over 2 minutes to show.We have thousands and sometimes even millions of records to report against.Queries have 4 and 5 table joins etc. We are using ASP.NET 2.0 in Visual Studio 2005 and Crystal Reports. Thanks
* We have a DAL that generates all SQL dynamically out of a nobject model. Standard very powerfull O/R mapper. * In the DAL, for CRUD operations, we generate the statements dynamically. As an example, let's take INSERT. * The insert is generated ONCE, with parameters, and cached. For every reuse, the parameters are replaced (in value), and the whole thing commited.
I see hte following negative: I can not easily batch multiple inserts. Parameters have to be unique per batch. So, if I want to batch two inserts, I need two sets of parameters.
Alternative:
Instead of generating the SQL with parameters, we generate the SQL as a string ready to be inserted for / with a String.Format, and then I encode the parameters and make one SQL String out of this. Now, please - don't say "sql injection", we are not that stupid, the layer handles this already, properly encoding all dangerous values.
With this approach, the SQL statement would be a string and not use any parameter. As a result, I could batch them up as much as I want (ok, up to a certain string size). I need to keep parameters around anyway (for blobs etc.), but most objects do not have blobs, and the SQL is prettty small. This small SQL could be batched significantly (100 statements per batch, propably mode) and be submitted to the database. As a result, the round trips to the databae would go down.
Now, my question is - which of the two approaches is more advisable, from a performance point of view? Again, stuff like SQL injection and ease of handling are totally irrelevant - the SQL never leaves the DAL and is generated in there, and we will go through a lot of complexitiy for higher performance.
Normally I would say batching should be better. SQL Server can auto-parameterize the statements (reusing the query plan), and / but the network round trips are the larger issue here.
I have a situation where currently, I check to see if a record exists and I do an update if it does, I do an INSERT if it does not. This happens about 200k times each night.
I'm trying to speed the application up that is doing this and I was wondering if it would be a better idea to just delete the records I'm about to recreate, before I start, then do an unconditional insert.
It would eliminate my IF EXISTS check but if an update is faster than an insert, it might even out.
Also, will SQL deal with the big delete and the data being recreated each night without too much bloating or fragmentation?
File I/O or SQL Database calls?Note that my SQL Server is on a separate server, so that cuts down on the web server resources needed per query, but it increases the time necessary for the query to happen.
Hi.We have stored procedure update specific table Each time it run it delete 5000- 6000 rows from table then insert 5000- 6000 rows with different information. It take up to 1 1/2 min execute.
1.Can force Sql server do not make entry for each insert and if yes would it increase speed of procedure ? 2. Is any other way increase speed of insert?
We intend to import up to 20+ GB of data into a database that will be a snapshot of month end data. The database tables must be dropped and then recreated just prior to importing the data. We are investigating BCP, DTS, and Bulk Insert. What is the fastest method for importing data given our approach?
We have been testing a new vendor_purchased application and are now running some month-end processes/jobs now. One of the processes that we kicked off stopped when a 2gb tempdb filled up with only 5mb or so remaini g. In reviewing what happened, I noticed that tempdb is in the process of clearing itself out slowly when we restarted the process. The database has 80% of its disk space remaining and the transaction log has also 75% remaining.
My questions are:
1) Although I don't think so, is there nay way of speeding up the process of tempdb clearing out the data in it?
2) We will need to examine what sql code the vendor has used that caused this to happen. Aside from group by and order by, if there are a lot of 'select into' code, what alternatives do we have?
Any information that can be provided will be fine. THanks in advance.
I need help makeing the following query run more efficently.
Code:
SELECT t1.ID,t1.firstName,t1.lastName,t1.address,t1.city,t1.state, t1.zip,t1.locationAddress,t1.locationCity,t1.locationState,t1.locationZip FROM Landlord_tbl t1 left outer join Mail_tbl t2 ON t2.potentialSitesID = t1.potentialSitesID WHERE t2.mailed_out_date is null and NOT(t1.firstName+t1.lastName) is Null GROUP BY t1.ID,t1.firstName,t1.lastName,t1.address,t1.city,t1.state, t1.zip,t1.locationAddress,t1.locationCity,t1.locationState,t1.locationZip ORDER BY t1.firstName, t1.lastName, t1.city, t1.state
We have a couple of 200GB databases that are recreated each night on a SQL2012 server connected to a disk array. The SQL disk is an auto-tiered combo of 10k and 7k drives in a RAID1 lun, and both data and log files are there.
Recently, some room has opened up on an older array that contains smaller, but many, 15k drives that I could use in a RAID1 config. Being that I'd like to split up the mdf and ldf files, which would you put onto the new (faster) disk?
EDIT: Add'l info: the only current performance issues I see in the SQL Log are FlushCache messages occuring throughout the night, when all activity happens for this DB. Things like this: "FlushCache: cleaned up 388690 bufs with 23474 writes in 409743 ms (avoided 179747 new dirty bufs) for db 47:0"
We have a financial accounting system that has been certified for SQL 2005. Before I upgrade, I'm wondering if I should expect better/faster performance from SQL 2005 vs. SQL 2000?
Anything I should lookout for when upgrading to SQL 2005?
Hi, when I execute the sql in DBTestArchive and then DBTest query analyzer , I found that the run time is shorter in DBTest. Can I make assumption that the query run time is shorter in DBTest if I select data from it and insert into another database? Thanks.
INSERT INTO DBTestArchive.dbo.tblVendorMasterArchive SELECT * FROM DBTest.dbo.tblVendorMaster
In a situation where you have two tables in a hierarchy like this:create table authors (authorid int identity (1,1))create table books (authorid int,bookid int identity (1,1))Is there any disadvantage to having the primary key and the clusteredindex as a compound key, like this:alter table books add constraint PK_books primary key clustered(authored, bookid)Normally, I would make bookid the key, but then I got to thinking, mostof the queries are going to be "select * from books where authorid =@@some_authorID"So, wouldn't a compound key and index make this a little faster?