hi, i'm using Access 2007 and i'm trying to join two selects and create two new columns[complete and not complete] where 'x' denotes a hit was made. i will use this later for grouping. here is my code so far. thanks.
SELECT tblOutlookTask.TaskSubject, tblOutlookTask.PercentComplete, tblOutlookTask.ID
FROM tblOutlookTask
WHERE (((tblOutlookTask.PercentComplete)=100))
SELECT tblOutlookTask.TaskSubject, tblOutlookTask.PercentComplete, tblOutlookTask.IDFROM tblOutlookTask
WHERE (((tblOutlookTask.PercentComplete)<>100))
I came across this structure today and haven't seen it before:
SELECT blablabla FROM T1 FULL OUTER JOIN T2 ON T1.Col1 = T2.Col1 AND T1.Col2 = T2.Col2 ON T3.Col1 = T1.Col1 AND T3.Col2 = T1.Col2 ON T4.Col1 = T1.Col1 AND T4.Col2 = T1.Col2
I remember when studying for exam 70-528 that I read a practice question that said it was better to return 2 select satements rather than a join on two tables as this would be more efficent. The question from MCPD Self Paced Training Kit: Designing and Developing Web Based Applications Using Microsoft .NET Framework (Certification Series): Designing and Developing ... Applications Using Microsoft .NET Framework was: You are an ASP.NET application developer. You are participating in the design of a Web retail application. The application accepts orders from the clients and stores them in a SQL Server 2005 database. Each order contains header information (order date, client information, and shipping address) and details information, which includes information about ordered products. Each detail line contains information about one ordered product and its quantity and price. All of the preceding information will be stored in two database tables: one table called Orders for the order header and another table, called OrderDetails, for order details in which each record in the orders table could have one or more related records in the OrderDetails table. When the user selects the specific order ID, the application is supposed to select all the information about the order from the database, and you need to create a stored procedure that allows selection of information about all the fields for the stored order from the database, based on a provided order ID value. Which of the following stored procedures do you recommend creating in a database to achieve the best performance on a database server side and reduce network traffic between application and database server?The answer was:CREATE PROCEDURE ORDERS_SELECT_ORDER @ORDER_ID INT AS SELECT * FROM Orders WHERE OrderID=@ORDER_IDSELECT * FROM OrderDetails WHERE OrderID=@ORDER_ID RETURNAs oppose to the traditional: CREATE PROCEDURE SP_SELECT_ORDER @ORDER_ID INT AS SELECT * FROM Orders LEFT JOIN OrderDetails ON Orders.OrderID = OrderDetails.OrderlD WHERE Orders.OrderID=@ORDER_IDThe explanation was: To achieve the best performance, avoid names of the stored procedures that start with sp_ prefix. If the name of the stored procedure starts with sp_ SQL Server always looks for the stored procedure in the master database first because all the system stored procedures start with the same prefix and are stored in a master database. It requires extra effort from the server to locate the stored procedure. This occurs even if you qualify the stored procedure with the database name. To avoid this issue, use a custom naming convention rather than the sp_ prefix. In addition, using a join inside of the stored procedure for the provided scenario will require an extra task from the server to join the data and produce a joined result. It will also lead to the redundant data for the header information that will be duplicated with each detail of the order and will increase traffic between the client and the server. Using only two separate SELECT SQL statements will allow you to avoid a join operation and reduce the size of the returned data to the client.Any one got any views on this as I have never actually seen this demostrated in a sample applcation. Is it a lot more efficent?ThanksScott
I have a database on SQL Sever 2005 SP1 against which a Publication has been defined and to which many servers (both Workgroup and Express editions - SP1) Subscribe to.
I would like to be able to distinguish between the Publisher and the Subscribers programmatically via T-SQL.
From reading BoL and various forums it appears that the IsPublished, IsMergePublished and IsSubscribed options of the DatabasePropertyEx function should give me this information.
However within all our tested environments, whilst the IsMergePublished option returns expected values. IsPublished and IsSubscribed both return 0 on all servers (the Publisher and Subscribers).
Is this a know issue and how can I rectify the problem or alternatively does anyone know of another method to distinguishing between the Publishers and Subscribers.
I'm an elementary C# developer.I find servers in local network using "SmoApplication.EnumAvailableSqlServers()" and fill a comboBox with them but I don't know how to distinguish SQLServers from SQLServer Express version. Can you help me please that what can I do? thanks a lot
Hi,I want to have all-in-one trigger, defined like this:CREATE TRIGGER MyInsertDeleteUpdateHandlerON MyTableFOR DELETE, INSERT, UPDATEASBEGIN(...)ENDNow, how can I tell why this trigger was fired (what event causedtrigger to be fired) - was it DELETE, INSERT or UPDATE?Is there something like this: @@event_type,so I could do for example IF (@@event_type = DELETE) (...)Of course I can create 3 triggers instead of 1, to be sure what eventfired my trigger.I can also count records in _deleted_, _inserted_ tables or to doJOINs with it. But, _inserted_ table is common for UPDATE and INSERTevents..Any suggestions?Thanks in advance.Hubert
I would like SQL Server 2000 to distinguish between uppercase and lowercase letters, but only within a single stored procedure. Also, at the end of the sp, I want the original collation to be restored. How will I implement this in my sp?
I need to be able to construct a statement that says something like:
If advertiserTYPE = FSBO Then INSERT INTO [COLUMN1] If advertiserTYPE = BROKER Then INSERT INTO [COLUMN2 If advertiserTYPE = DEVELOPER Then INSERT INTO [COLUMN3]
I have this sproc. ALTER PROCEDURE dbo.cis_UpdateCourseUserWithGrade @Grade nvarchar, @UaaStudentId nchar, @CourseId int AS UPDATE cis_CourseUser SET Grade =@Grade WHERE UaaStudentId = @UaaStudentId AND WHERE CourseID = @CourseId RETURN the ANE WHERE isn't much liked. How can I update the field with the grade where both of the conditions are true? I'm getting an invalid syntax near @UaaStudentId
I've seen lots of entries recommending the use of ISNULL in SQL WHERE clauses, e.g. in a search sproc where users can enter some or all parameters to search a table. Previously I would have used something like:SELECT * FROM MyTableWHERE (FName = @fname OR @fname IS NULL) AND(MName = @mname OR @mname IS NULL) AND(LName = @lname OR @lname IS NULL)So using the neat ISNULL syntax it could be updated to:SELECT * FROM MyTableWHERE (FName = ISNULL(@fname, FName)) AND(MName = ISNULL(@mname, MName)) AND(LName = ISNULL(@lname, LName))Having played around with this I stumbled upon a problem. If one of the fields, e.g. MName, is NULL then that clause will return false since MName = NULL isn't true and you have to use MName IS NULL. Did I miss all the caveats with using ISNULL in this way on fields that can contain NULL or have I missed something else?
So how do you do dynamic WHERE clauses with asp.net? I have a QueryString parameter with a value that will look like this. |1| or |1||2| or |1||2||3| or etc. For each number in the value I want to add an EXISTS function to the WHERE clause. I have accomplished this in T-SQL. But I can't figure out how to bind asp.net control to it. It gives me 'Invalid syntax near the keyword 'ORDER'. I know this has something to do with .net not being able to exaluate the schema properly. So I tried Linq to Sql. Well, so happens that my WHERE clause is pointing to the existance of records in another table which the default install of Linq to Sql doesn't seam to support. I came across this post (http://forums.microsoft.com/MSDN/ShowPost.aspx?PostID=2979081&SiteID=1) that explains how to dynamically query across multiple tables. I tried this and got errors when trying to plug in the additional code. And this still does not let me dynamically add where clauses. I tried the who Predicate thing and ran into the same problem where I can't reference the existance of values in foreign tables. Here is my sproc. Can anyone tell me how to get this into a asp.net environment. I also do NOT want to auto-generate fields in my GridView.ALTER PROCEDURE dbo.Item_SelectByFilters @Item_Category_Id int = NULL, @ItemSpecificValueIdList varchar(100) = NULL, @debug bit = 0 AS DECLARE @sql nvarchar(4000), @paramlist nvarchar(4000) SELECT @sql = 'SELECT Item_Id, Number, Primary_Item_Category_Id, Secondary_Item_Category_Id, Engineering_Document_Id, CategoryIdList FROM Item WHERE (1 = 1)' IF @Item_Category_Id IS NOT NULL SELECT @sql = @sql + ' AND Item.CategoryIdList Like ''|'' + CONVERT(varchar, (@xItem_Category_Id) + ''|''' IF @ItemSpecificValueIdList IS NOT NULL BEGIN WHILE @ItemSpecificValueIdList <> '' BEGIN DECLARE @StartLocation int SELECT @StartLocation = CHARINDEX('-', @ItemSpecificValueIdList, 1) IF @StartLocation <> 0 BEGIN DECLARE @EndLocation int SELECT @EndLocation = CHARINDEX('|', @ItemSpecificValueIdList, @StartLocation + 1) IF @EndLocation <> 0 BEGIN DECLARE @ValueLength int SELECT @ValueLength = @EndLocation - @StartLocation - 1 DECLARE @ValueBetween varchar(10) SELECT @ValueBetween = SUBSTRING(@ItemSpecificValueIdList, @StartLocation + 1, @ValueLength) IF @ValueBetween IS NOT NULL SELECT @sql = @sql + ' AND EXISTS(SELECT Item_Specific_Value_Id FROM Item_Specific_Value WHERE Item_Id = Item.Item_Id AND Item_Specific_Value.Item_Specific_Value_Id = ' + @ValueBetween + ')' END END SET @ItemSpecificValueIdList = SUBSTRING(@ItemSpecificValueIdList, @EndLocation + 1, len(@ItemSpecificValueIdList) - @EndLocation) END END SELECT @sql = @sql + ' ORDER BY Item.Number' IF @debug = 1 PRINT @sql SELECT @paramlist = '@xItem_Category_Id int' EXEC sp_executesql @sql, @paramlist, @Item_Category_Id
Here is another project that seams to be the same as my previous post but a little simpler. http://forums.asp.net/t/1263330.aspx My Item_Category table is self referencing. I have a list of category Ids in a particular order representing the tree of categories from the current category to the top. For example: 1 - Fasteners, 4 - Screws, 12 - Sheet Metal Screws or the reverse. I need to translate this so I can bind a cookie crumb type control to it. ie. Fasteners > Screws > Sheet Metal ScrewsI originally accomplished this like so; Dim q = From ic In itemdc.Item_Category _Where valueId.Contains(ic.Item_Category_Id) _ Select Id = ic.Item_Category_Id, ic.Name But found that this does not order them properly. Remember that the list of Ids is in a particular order. I need a query to result to the below. I was using a horizontal DataList control as the cookie crumb control. Again, I am trying to do this with Linq to Sql without using Stored Procedures even though I am the DBA as well. Or is there some other better way to do this?
Id Name Sort
1 Fastener 1
4 Screws 2
12 Sheet Metal Screws 3 Finally, as I mentioned in my previous post, I accomplished this concept with a stored procedure but then the asp.net controls could not recognize the schema, hence I could not bind the controls to the data source control.
I have an existing SPROC which works the way it should do and was difficult to construct. But now I need to add 1 more condition and hopefully that should be it however I am struggling how to do this. Here is a snip of the existing SPROC.
quote: IF @columnName = 'Rating - Fire' OR @columnName = 'Rating - PPE' OR @columnName = 'Rating - Reactivity' OR @columnName = 'Rating - Health' BEGIN SELECT @totalRecords = (SELECT COUNT(p.[SID]) FROM S_Summary p INNER JOIN S_Detail detail ON detail.SID = p.SID WHERE CASE @columnName
[code]...
so this works fine but now I need to add 1 more thing..if a new parameter is supplied (lets call it @stringBranch), then I want to join another table and also match the param value to a field in that table along with any existing WHERE conditions being applied to this:
quote: WHERE (@columnName IS NULL AND @columnValue IS NULL) OR CASE @columnName WHEN 'Rating - Fire' THEN detail.F WHEN 'Rating - PPE' THEN detail.P
I'm trying to use multiple where clauses but its not working. I want it to look something like this:
INSERT INTO [USCondex_Development].[dbo].[miamiheraldExceptions]([InvalidEmails], [InvalidAdPrintID], [InvalidPropertyStreetAddress], [InvalidPropertyPrice]) SELECT [AdvertiserEmail],[AdPrintId],[AdvertiserAddress], [PropertyPrice] FROM [Development].[dbo].[table2] WHERE advertiseremail is NULL and WHERE adPrintID is NULL and WHERE firstinsertdate is NOT NULL and WHERE propertystreetaddress is NOT NULL and WHERE propertyprice < 100
I am trying to wrap my WHERE clause with an IF or a CASE but cannot seem to get it to work. This is what I am trying:
WHERE CASE WHEN EVENT_TYPE='d' THEN (link_inc.incident_id = 10000005) AND (B.incident_id <> 10000005) AND link_rsn.link_rsn_sc = 'CHANGE' AND B.incident_id > 10000000 ELSE (link_inc.incident_id = 10000005) AND (B.incident_id <> 10000005) AND link_rsn.link_rsn_sc = 'CHANGE' AND B.incident_id > 10000000 AND act_type.act_type_sc <> 'CLOSURE' END ORDER BY B.incident_id DESC, act_reg.act_reg_id DESC
Basically I want to run a different WHERE clause based on a value (EVENT_TYPE). The error message I am getting is: Incorrect syntax near '='.
I have a table, basically consisting of products and their prices. I want to select some products, then sort them by price in ascending order BUT putting prices of zero at the bottom. (e.g. 5.99, 8.99, 10.99, 0.00, 0.00)I thought I'd be able to do something like:ORDER BY (price != 0), pricethinking that it would sort rows according to whether the condition was true or not, and then by price, but MSSQL doesn't seem to allow this. should this work, or is there another way around this? One solution would be to load the values into a table object and sort them using that, but I'd rather do all of this in SQL if possible, for speed.any suggestions?thanks!
If I have a table called "content_hits_tbl" and want to pull information, can't i write something like this:
SELECT COUNT(visitor_id) AS HITS, COUNT(DISTINCT visitor_id) AS VISITORS, COUNT(DISTINCT visitor_id) WHERE visit_type = 0 AS NEW, COUNT(DISTINCT visitor_id) WHERE visit_type = 1 AS RETURNING FROM content_hits_tbl
Can't you have multiple WHERE clauses in the SELECT statement?
Any suggestions would be great. I have been wrestling with it and SQL queries arent my strong area ...
I have a Master table with a OrderNbr which is also contained in the Detail table.
It's a 1 to Many relationship, respectively.
I want to update the MASTER.FinalizedDate using a "select top 1 FinalizedDate order by FinalizedDate DESC" from the Detail table but the clause is ALL the Status have to be "F". So OrderNbr 12345 should not get updated because it contains a 'O'. OrderNbr 67899 should get updated in the Master table to 2/26/2013 because all have a 'F' and the last date to post is the official finalized date.
Here is what I came up with.......so far, but not sure how to work in the Status piece on 1 to M.
The rub here is that even if one row has the 'O' status I want to ignore the update. If all have the 'F' then I want the udpate to happen.
Update MASTER Set FinalizedDate = (select top 1 d.FinazliedDate from Detail d where m.OrderNbr = d.OrderNbr and d.Status not in ('O') Order by FinalizedDate DESC) From MASTER m
How do I not include all 3 rows for OrderNbr 12345 because one row has the Status "O" in the DETAIL table?
Here are the table looks........
MASTER OrderNbr Ytotals Ztotals Xtotals Finalized Date 12345$1,500$1,500$1,200 67899$1,200$1,100$900
In a UDF, how is the best way to extend a query with additional clauses based on expressions? The user input here is used to refine the basic query by introducing additional clauses. Is there something like the following?
-- The basic query SELECT column FROM table WHERE clause
-- Additional clause, only appended to query -- if expression evaluates to true IF @parameter <> default_value BEGIN AND additional_clause END
Right now I'm using CASE like the following, but it necessarily makes the query longer. Is there a more efficient way?
-- The basic query
SELECT column FROM table WHERE clause
-- Additional clause, should only effect result -- set when the parameter is not default_value AND table.column = CASE @parameter WHEN default_value THEN -- identity, table.column=table.column, -- should have no effect other than just a long query table.column ELSE @parameter END
Hi I'm not sure if I have stated my subject line correctly for what I want to achieve, but I will attempt to explain it below.
In addition to what I have in my script below, I also need to include the following clauses:
1. where TransPerPaySequence.FinancialYTDCode like '2007', and e.EmployeeStatusCode like 'CASUAL' and p.PositionGroupCode like 'AC', then instead of using the divisor of 72, it needs to be 35; and
2. where TransPerPaySequence.FinancialYTDCode like '2008', and e.EmployeeStatusCode like 'CASUAL' and p.PositionGroupCode like 'AC', then instead of using the divisor of 72 or 35, it needs to be 31.
I would really appreciate any assistance that can be provided.
Thanks
SELECT DISTINCT pc.PositionClassificationCode, pc.Description AS positionclass, pg.PositionGroupCode, pg.Description AS positiongroup, p.Description AS position, e.PreferredName + ' ' + e.LastName AS employeename, SUM(ha.Quantity) / ((CASE p2.PositionGroupCode WHEN 'AC' THEN 72 WHEN 'AL' THEN 75 WHEN 'EX' THEN 80 WHEN 'MG' THEN 80 WHEN 'SM' THEN 80 END) * (SELECT COUNT(DISTINCT PaySequence) AS Expr1 FROM TransPerPaySequence WHERE (PayPeriodCode LIKE 'EIT') AND (Closed = '1') AND (Description LIKE 'St%'))) AS FTE, (SELECT COUNT(DISTINCT PaySequence) AS Expr1 FROM TransPerPaySequence AS TransPerPaySequence_1 WHERE (PayPeriodCode LIKE 'EIT') AND (Closed = '1') AND (Description LIKE 'St%')) AS payseq FROM HistoricalAllowance AS ha LEFT OUTER JOIN Position AS p ON ha.PositionCode = p.PositionCode LEFT OUTER JOIN PositionGroup AS pg ON p.PositionGroupCode = pg.PositionGroupCode LEFT OUTER JOIN PositionClassification AS pc ON p.PositionClassificationCode = pc.PositionClassificationCode LEFT OUTER JOIN WAP ON ha.WAPCode = WAP.WAPCode LEFT OUTER JOIN Employee AS e ON ha.EmployeeCode = e.EmployeeCode LEFT OUTER JOIN Position AS p2 ON e.PositionCode = p2.PositionCode LEFT OUTER JOIN TransPerPaySequence AS tpps ON ha.PaySequence = tpps.PaySequence WHERE (e.EmployeeCode IN ('83', '739')) AND (ha.AllowanceCode IN ('005', '201', '203', '101')) AND (tpps.FinancialYTDCode LIKE '2007%') GROUP BY pc.PositionClassificationCode, pg.PositionGroupCode, pc.Description, pg.Description, p.Description, e.PreferredName, e.LastName, p2.PositionGroupCode
Hi, I hope some one can help me. I have a stored procedure (Microsoft SQL 2005 Express Edition) that I want users to be able to dynamically set the, group by, order by (@orderby) and where clause (@where). I have managed to get the group by to work but can't seem to get the where and order by to work. Here's my stored procedure. Any idea how this can be done? ALTER PROCEDURE [dbo].[sp_aggregate] -- Add the parameters for the stored procedure here @finfileid int, @phaseid int, @supplierid int, @measurementid int, @roleid int, @groupby int, @orderby int, @where int AS BEGIN -- SET NOCOUNT ON added to prevent extra result sets from -- interfering with SELECT statements. SET NOCOUNT ON; -- Insert statements for procedure here SELECT MAX(ProjectFinFileMonthItems.ProjFinFileMonthItemsMonthId) AS ProjFinFileMonthItemsMonthId, SUM(ProjectFinFileMonthItems.ProjFinFileMonthItemsValue * ProjectFinFileMonthItems.ProjFinFileMonthItemsRate * ProjectFinFileMonthItems.ProjFinFileMonthItemsAvail / 100) AS total, MAX(ProjectFinFileItems.ProjPhaseId) AS phaseid, MAX(ProjectFinFileMonthItems.ProjDeliveId) AS deliveid, MAX(ProjectFinFileMonthItems.SupplierId) AS supplierid, MAX(ProjectFinFileMonthItems.ProjFinFileItemsId) AS ProjFinFileItemsId, MAX(ProjectFinFileMonthItems.ProjFinFileMonthItemsId) AS ProjFinFileMonthItemsId, MAX(ProjectFinFileMonthItems.ProjDeliveId) AS ProjDeliveId, MAX(ProjectPhases.ProjectPhaseName) AS ProjectPhaseName, MAX(Suppliers.SupplierName) AS SupplierName, MAX(ProjectFinFileMonthItems.RoleId) AS RoleId, MAX(Measurements.MeasurementName) AS MeasurementName, MAX(ProjectFinFileMonthItems.MeasurementId) AS MeasurementId, MAX(ProjectFinFileMonthItems.FinDataTypeId) AS FinDataTypeId, MAX(FinDataTypes.FinDataTypeName) AS FinDataTypeName, max(ProjectFinFileItems.FinFileId) as finfileid FROM ProjectFinFileItems INNER JOIN ProjectFinFileMonthItems ON ProjectFinFileItems.ProjFinFileItemsId = ProjectFinFileMonthItems.ProjFinFileItemsId LEFT OUTER JOIN FinDataTypes ON ProjectFinFileMonthItems.FinDataTypeId = FinDataTypes.FinDataTypeId LEFT OUTER JOIN Measurements ON ProjectFinFileMonthItems.MeasurementId = Measurements.MeasurementId LEFT OUTER JOIN Roles ON ProjectFinFileMonthItems.RoleId = Roles.RoleId LEFT OUTER JOIN ProjectPhases ON ProjectFinFileItems.ProjPhaseId = ProjectPhases.ProjectPhaseId LEFT OUTER JOIN Suppliers ON ProjectFinFileMonthItems.SupplierId = Suppliers.SupplierId /*dynamic where clause needs to go here */ /*dynamic group by clause */ GROUP BY CASE when @groupby=1 then ProjectFinFileItems.projphaseid --phaseid when @groupby=2 then ProjectFinFileMonthItems.supplierid -- supplierid when @groupby=3 then ProjectFinFileMonthItems.measurementid -- measurment when @groupby=4 then ProjectFinFileMonthItems.roleid --role else ProjectFinFileMonthItems.ProjFinFileMonthItemsId END /*dynamic order clause needs to go here */ END cheers Mark :)
I have a web page which passes back parameters to a stored procedure. From the web page the user selects different clauses for the 'where' criteria. Based upon the number of clause items in the parameters sent back, a select statement is built and executed. In the stored procedure I have many if statements to chose the correct sql statement. As the no of clauses in the where statement can vary, it can become messy script. Has anyone dealt with this scenario. What is the best strategy ?
A simple illustation of this is as follows A statement with two clauses :-
Select * from Sales where user = 'John' and country = 'England' A statement with three clauses :- Select * from Sales where user = 'John' and country = 'England' and County = 'Staffordshire'
The stored procedure would except three parameters and would build a string based on the number of actual where clases sent back
I have a complex SQL query against multiple tables that ideally would be best used as a stored procedure.
I know of and have read about creating named parameters and setting the parameters' values in ASP just before issuing an execute command via the Command object. What I'd like to do is assign the text of the filter and sort clauses to the parameters. Not the values, but the entire string, such as
"WHERE lastname LIKE 'A%' "
"ORDER BY lastname "
I need to assign the entire clause because the query may or may not use a particular clause.
I have experimented. However SQL treats the parameter as a literal string rather than part of the SQL query itself. Is there another technique that I may use to accomplish my goal?
Thanks in advance for your opinion, suggestion, criticism, etc.
I'm self-taught on T-SQL so forgive me if this is a dumb question. If I have a stored procedure containing the following...
IF [condition] SELECT ... ELSE SELECT ...
Would it be more efficient/proper/etc. to break those two SELECTs out into their own SPs and execute them from the IF statement?
IF [condition] EXEC [SP] ELSE EXEC [SP]
I've got a number of pretty lengthy SPs that I think are pretty tight but if I can make them more efficient by breaking them down into smaller tasks, I'd rather do that.
What is your opinion using Sub Selects versus Joining Tables?
SELECT (SELECT COMPANY_NAME FROM COMPANIES WHERE COMPANY_ID = c.COMPANY_ID) AS 'Company Name'
FROM COMPANY_ORDERS c
vs
SELECT co.COMPANY_NAME AS 'Company Name' FROM COMPANIES c, COMPANY co WHERE c.COMPANY_ID = co.COMPANY_ID
I'm not having any problems, just curious what everyone is practicing and is "system-actically" more resourceful. I personally like using Sub Selects (or Sub Queries) so this way I wouldn't have to deal with complex joins and where conditions, especially when your just dealing with a releational table.
I'm having problems optimizing a sql select statement that uses a LIKE statement coupled with an OR clause. For simplicity sake, I'll demonstrate this with a scaled down example:
CompanyAddressAssoc is the many-to-many associative table for Company and Address. A search query is required that, given a search string ( i.e. 'TEST' ), return all Company -> Address records where either the CompanyName or AddressName starts with the parameter:
Select c.CompanyID, c.CompanyName, a.AddressName
FROM Company c
LEFT OUTER JOIN CompanyAddressAssoc caa ON caa.CompanyID = c.CompanyID
LEFT OUTER JOIN Address a ON a.AddressID = caa.AddressID
WHERE ((c.CompanyName LIKE 'TEST%') OR (a.AddressName LIKE 'TEST%))
There are proper indexes on all tables. The execution plan creates a hash table on one LIKE query, then meshes in the other LIKE query. This takes a very long time to do, given a dataset of 500,000+ records in Company and Address.
Is there any way to optimize this query, or is it a problem with the base table implementation?
