asked on

Need help updating stored procedure to run via vb.net

Hi,
I need to run a stored procedure from my Windows Application. I added it to my dataset and when I try to run it it gives me an error because it has temporary tables in it. I tried using table variables instead but then the stored procedure takes too long to run. The original procedure runs in 15 seconds where the new one I created runs past 8 minutes.

How can I update my stored procedure so I can run it from my Windows application without sacrificing performance? I've posted my original sp code and the new one I created

Original Stored Procedure with Temp tables

BEGIN
	SET NOCOUNT ON

	CREATE TABLE #ClaimCountsByAge
	(
		AgeRange VARCHAR(25)
		,NewClaims INT
		,NewClaimsPaidSum DECIMAL(20,4)
		,FollowUpClaims INT
		,FollowUpClaimsPaidSum DECIMAL(20,4)
		,ClosedClaims INT
		,ClosedClaimsPaidSum DECIMAL(20,4)
	)

	DECLARE @ValueRange TABLE
	(
		AgeRange VARCHAR(20)
		,StartAge INT
		,EndAge INT
		,PRIMARY KEY CLUSTERED (AgeRange, StartAge, EndAge)
	)

	INSERT INTO @ValueRange VALUES ('0-90',	0, 90)
	INSERT INTO @ValueRange VALUES ('91-180', 91, 180)
	INSERT INTO @ValueRange VALUES ('181-270', 181, 270)
	INSERT INTO @ValueRange VALUES ('271-360', 271, 360)
	INSERT INTO @ValueRange VALUES ('361-450', 361, 450)
	INSERT INTO @ValueRange VALUES ('451-540', 451, 540)
	INSERT INTO @ValueRange VALUES ('541-630', 541, 630)
	INSERT INTO @ValueRange VALUES ('631-720', 631, 720)
	INSERT INTO @ValueRange VALUES ('721-810', 721, 810)
	INSERT INTO @ValueRange VALUES ('811-900', 811, 900)
	INSERT INTO @ValueRange VALUES ('901-990', 901, 990)
	INSERT INTO @ValueRange VALUES ('991-1080', 991, 1080)
	INSERT INTO @ValueRange VALUES ('1081-1170', 1081, 1170)
	INSERT INTO @ValueRange VALUES ('1170+', 1171, 10000000)	

	SELECT 
		cl.tarsClaimID
		,cl.ICN
		,cl.TMHPPaid
		,cl.FollowUpUser
		,cl.CompleteUser
		,cl.CompleteDate
	INTO
		#ClaimsInfo
	FROM
		dbo.Claims cl
	WHERE
		( (cl.FollowUpUser IS NULL
			AND cl.CompleteUser IS NULL
			AND cl.CompleteDate IS NULL)
		 OR (cl.FollowUpUser IS NOT NULL
			AND cl.FollowUpUser NOT IN ('hms', 'system'))
		 OR (cl.CompleteUser IS NOT NULL
			AND cl.CompleteUser NOT IN ('hms', 'system')))

	CREATE CLUSTERED INDEX idx_ClaimsInfo__tarsClaimID__ICN ON #ClaimsInfo(tarsClaimID, ICN)
	CREATE INDEX idx_ClaimsInfo__User ON #ClaimsInfo(FollowUpUser, CompleteUser) INCLUDE (TMHPPaid)

	SELECT
		x.ICN
		,(CASE 
			WHEN x.CompleteDate IS NULL THEN DATEDIFF(d, x.LDOS, GETDATE())
			ELSE DATEDIFF(d, x.LDOS, x.CompleteDate)
		END) AS ClaimAge
	INTO
		#ClaimsByAge
	FROM
		(SELECT  
			ci.ICN
			,max(dtl.toDOS) AS LDOS
			,max(ci.CompleteDate) AS CompleteDate
		FROM 
			#ClaimsInfo ci
			INNER JOIN dbo.Detail dtl
				ON ci.tarsClaimID = dtl.tarsClaimID
		GROUP BY
			ci.ICN) x

	CREATE CLUSTERED INDEX idx_ClaimsByAge__ICN__ClaimAge ON #ClaimsByAge(ICN, ClaimAge)

-- New/Remaining Claim COunt
	INSERT INTO #ClaimCountsByAge 
		(
			AgeRange 
			,NewClaims
			,NewClaimsPaidSum 
		)
		SELECT 
			x.AgeRange
			,COUNT(x.ICN)
			,SUM(x.TMHPPaid)
		FROM 
			(SELECT 
				vr.AgeRange
				,ci.ICN
				,max(TMHPPaid) as TMHPPaid
			FROM 
				#ClaimsInfo ci
				INNER JOIN #ClaimsByAge ca 
					ON ca.ICN = ci.ICN
				INNER JOIN @ValueRange vr
					ON ca.ClaimAge BETWEEN vr.StartAge AND vr.EndAge
			WHERE 
				ci.CompleteUser IS NULL 
				AND ci.CompleteDate IS NULL --not complete
				AND ci.FollowUpUser IS NULL -- not currently being worked
				-- no associated payment
				AND NOT EXISTS (SELECT 
									1
								FROM 
									dbo.Finr226c f WITH(NOLOCK) 
								WHERE 
									f.XRef = ci.ICN) 
			GROUP BY 
				vr.AgeRange
				,ci.ICN) x
		GROUP BY
			x.AgeRange

-- Follow-Up Claim COunt
	INSERT INTO #ClaimCountsByAge 
		(
			AgeRange 
			,FollowUpClaims
			,FollowUpClaimsPaidSum 
		)
		SELECT 
			x.AgeRange
			,COUNT(x.ICN)
			,SUM(x.TMHPPaid)
		FROM 
			(SELECT 
				vr.AgeRange
				,ci.ICN
				,max(TMHPPaid) as TMHPPaid
			FROM 
				#ClaimsInfo ci
				INNER JOIN #ClaimsByAge ca 
					ON ca.ICN = ci.ICN
				INNER JOIN @ValueRange vr
					ON ca.ClaimAge BETWEEN vr.StartAge AND vr.EndAge
			WHERE 
				ci.CompleteUser IS NULL 
				AND ci.CompleteDate IS NULL --not complete
				AND ci.FollowUpUser IS NOT NULL 
				AND ci.FollowUpUser NOT IN ('hms', 'system') 
			GROUP BY 
				vr.AgeRange
				,ci.ICN) x
		GROUP BY
			x.AgeRange

-- Closed Claim COunt
	INSERT INTO #ClaimCountsByAge 
		(
			AgeRange 
			,ClosedClaims
			,ClosedClaimsPaidSum 
		)
		SELECT 
			x.AgeRange
			,COUNT(x.ICN)
			,SUM(x.TMHPPaid)
		FROM 
			(SELECT 
				vr.AgeRange
				,ci.ICN
				,max(TMHPPaid) as TMHPPaid
			FROM 
				#ClaimsInfo ci
				INNER JOIN #ClaimsByAge ca 
					ON ca.ICN = ci.ICN
				INNER JOIN @ValueRange vr
					ON ca.ClaimAge BETWEEN vr.StartAge AND vr.EndAge
			WHERE 
				ci.CompleteDate IS NOT NULL 
				AND ci.CompleteUser NOT IN ('hms', 'system') 
			GROUP BY 
				vr.AgeRange
				,ci.ICN) x
		GROUP BY
			x.AgeRange

	SELECT 
		vr.AgeRange
		,vr.StartAge 
		,MAX(cl.NewClaims) AS NewClaimsCount
		,MAX(cl.NewClaimsPaidSum) AS NewClaimsPaidSum
		,MAX(cl.FollowUpClaims) AS FollowUpClaimsCount
		,MAX(cl.FollowUpClaimsPaidSum) AS FollowUpClaimsPaidSum
		,MAX(cl.ClosedClaims) AS ClosedClaimsCount
		,MAX(cl.ClosedClaimsPaidSum) AS ClosedClaimsPaidSum
	FROM 
		@ValueRange vr
		LEFT OUTER JOIN #ClaimCountsByAge cl
			ON vr.AgeRange = cl.AgeRange
	GROUP BY
		vr.AgeRange
		,vr.StartAge 


	DROP TABLE #ClaimsInfo
	DROP TABLE #ClaimsByAge
	DROP TABLE #ClaimCountsByAge

END

Open in new window

New stored procedure using table variables

BEGIN
	SET NOCOUNT ON

	Declare @ClaimCountsByAge table
	(
		AgeRange VARCHAR(25)
		,NewClaims INT
		,NewClaimsPaidSum DECIMAL(20,4)
		,FollowUpClaims INT
		,FollowUpClaimsPaidSum DECIMAL(20,4)
		,ClosedClaims INT
		,ClosedClaimsPaidSum DECIMAL(20,4)
	)

	DECLARE @ValueRange TABLE
	(
		AgeRange VARCHAR(20)
		,StartAge INT
		,EndAge INT
		,PRIMARY KEY CLUSTERED (AgeRange, StartAge, EndAge)
	)

	INSERT INTO @ValueRange VALUES ('0-90',	0, 90)
	INSERT INTO @ValueRange VALUES ('91-180', 91, 180)
	INSERT INTO @ValueRange VALUES ('181-270', 181, 270)
	INSERT INTO @ValueRange VALUES ('271-360', 271, 360)
	INSERT INTO @ValueRange VALUES ('361-450', 361, 450)
	INSERT INTO @ValueRange VALUES ('451-540', 451, 540)
	INSERT INTO @ValueRange VALUES ('541-630', 541, 630)
	INSERT INTO @ValueRange VALUES ('631-720', 631, 720)
	INSERT INTO @ValueRange VALUES ('721-810', 721, 810)
	INSERT INTO @ValueRange VALUES ('811-900', 811, 900)
	INSERT INTO @ValueRange VALUES ('901-990', 901, 990)
	INSERT INTO @ValueRange VALUES ('991-1080', 991, 1080)
	INSERT INTO @ValueRange VALUES ('1081-1170', 1081, 1170)
	INSERT INTO @ValueRange VALUES ('1170+', 1171, 10000000)	

	
	
	Declare @ClaimsInfo table
	(tarsClaimID varchar(14) not null,
	ICN varchar(24),
	TMHPPaid money,
	Followupuser varchar(25),
	CompleteUser varchar(25),
	Completedate datetime,
	Primary key clustered(tarsClaimID, ICN))
	
	
	Insert into @ClaimsInfo
	
	
	SELECT 
		cl.tarsClaimID
		,cl.ICN
		,cl.TMHPPaid
		,cl.FollowUpUser
		,cl.CompleteUser
		,cl.CompleteDate

	FROM
		dbo.Claims cl
	WHERE
		( (cl.FollowUpUser IS NULL
			AND cl.CompleteUser IS NULL
			AND cl.CompleteDate IS NULL)
		 OR (cl.FollowUpUser IS NOT NULL
			AND cl.FollowUpUser NOT IN ('hms', 'system'))
		 OR (cl.CompleteUser IS NOT NULL
			AND cl.CompleteUser NOT IN ('hms', 'system')))



Declare @ClaimsbyAge table
(ICN varchar(24),
Claimage int)


Insert into @ClaimsbyAge
	SELECT
		x.ICN
		,(CASE 
			WHEN x.CompleteDate IS NULL THEN DATEDIFF(d, x.LDOS, GETDATE())
			ELSE DATEDIFF(d, x.LDOS, x.CompleteDate)
		END) AS ClaimAge

	FROM
		(SELECT  
			ci.ICN
			,max(dtl.toDOS) AS LDOS
			,max(ci.CompleteDate) AS CompleteDate
		FROM 
			@ClaimsInfo ci
			INNER JOIN dbo.Detail dtl
				ON ci.tarsClaimID = dtl.tarsClaimID
		GROUP BY
			ci.ICN) x


-- New/Remaining Claim COunt
	INSERT INTO @ClaimCountsByAge 
		(
			AgeRange 
			,NewClaims
			,NewClaimsPaidSum 
		)
		SELECT 
			x.AgeRange
			,COUNT(x.ICN)
			,SUM(x.TMHPPaid)
		FROM 
			(SELECT 
				vr.AgeRange
				,ci.ICN
				,max(TMHPPaid) as TMHPPaid
			FROM 
				@ClaimsInfo ci
				INNER JOIN @ClaimsByAge ca 
					ON ca.ICN = ci.ICN
				INNER JOIN @ValueRange vr
					ON ca.ClaimAge BETWEEN vr.StartAge AND vr.EndAge
			WHERE 
				ci.CompleteUser IS NULL 
				AND ci.CompleteDate IS NULL --not complete
				AND ci.FollowUpUser IS NULL -- not currently being worked
				-- no associated payment
				AND NOT EXISTS (SELECT 
									1
								FROM 
									dbo.Finr226c f WITH(NOLOCK) 
								WHERE 
									f.XRef = ci.ICN) 
			GROUP BY 
				vr.AgeRange
				,ci.ICN) x
		GROUP BY
			x.AgeRange

-- Follow-Up Claim COunt
	INSERT INTO @ClaimCountsByAge 
		(
			AgeRange 
			,FollowUpClaims
			,FollowUpClaimsPaidSum 
		)
		SELECT 
			x.AgeRange
			,COUNT(x.ICN)
			,SUM(x.TMHPPaid)
		FROM 
			(SELECT 
				vr.AgeRange
				,ci.ICN
				,max(TMHPPaid) as TMHPPaid
			FROM 
				@ClaimsInfo ci
				INNER JOIN @ClaimsByAge ca 
					ON ca.ICN = ci.ICN
				INNER JOIN @ValueRange vr
					ON ca.ClaimAge BETWEEN vr.StartAge AND vr.EndAge
			WHERE 
				ci.CompleteUser IS NULL 
				AND ci.CompleteDate IS NULL --not complete
				AND ci.FollowUpUser IS NOT NULL 
				AND ci.FollowUpUser NOT IN ('hms', 'system') 
			GROUP BY 
				vr.AgeRange
				,ci.ICN) x
		GROUP BY
			x.AgeRange

-- Closed Claim COunt
	INSERT INTO @ClaimCountsByAge 
		(
			AgeRange 
			,ClosedClaims
			,ClosedClaimsPaidSum 
		)
		SELECT 
			x.AgeRange
			,COUNT(x.ICN)
			,SUM(x.TMHPPaid)
		FROM 
			(SELECT 
				vr.AgeRange
				,ci.ICN
				,max(TMHPPaid) as TMHPPaid
			FROM 
				@ClaimsInfo ci
				INNER JOIN @ClaimsByAge ca 
					ON ca.ICN = ci.ICN
				INNER JOIN @ValueRange vr
					ON ca.ClaimAge BETWEEN vr.StartAge AND vr.EndAge
			WHERE 
				ci.CompleteDate IS NOT NULL 
				AND ci.CompleteUser NOT IN ('hms', 'system') 
			GROUP BY 
				vr.AgeRange
				,ci.ICN) x
		GROUP BY
			x.AgeRange

	SELECT 
		vr.AgeRange as [Age(Days)]
		,vr.startage		
		,MAX(cl.NewClaims) AS [New Claims]
		,MAX(cl.NewClaimsPaidSum) AS [New Claims Paid Sum]
		,MAX(cl.FollowUpClaims) AS [Follow-Up]
		,MAX(cl.FollowUpClaimsPaidSum) AS [Follow-Up Paid Sum]
		,MAX(cl.ClosedClaims) AS [Closed Count]
		,MAX(cl.ClosedClaimsPaidSum) AS [Closed Paid Sum]
	FROM 
		@ValueRange vr
		LEFT OUTER JOIN @ClaimCountsByAge cl
			ON vr.AgeRange = cl.AgeRange
	GROUP BY
		vr.AgeRange
		,vr.StartAge 



END

Open in new window

ASKER CERTIFIED SOLUTION

Russell Fox

membership

This solution is only available to members.

To access this solution, you must be a member of Experts Exchange.

Start Free Trial

David Johnson, CD

I added it to my dataset and when I try to run it it gives me an error because it has temporary tables in it.

Under which account and which security permissions are you running the stored procedure? You may just have to alter your database configuration to run as dbo or sysadmin

imstac73

ASKER

I tried changing them to normal tables. It works fine in Management Studio but when I add it to my dataset in my Windows application it gives me an invalid object name on all tables I am creating in the stored procedure.

Russell Fox

Sounds like a problem with the credentials on the account that's running the Windows app. If that account doesn't have admin rights on the server, you may need to grant it rights on the new tables:

GRANT ALL ON dbo.Temp_ClaimCountsByAge TO YourWindowsUserAccount;

Open in new window

imstac73

ASKER

The Windows application is ran by whoever is logged into it. So when I do the testing it is my account that is running it. I am also using the same account in Management Studio. So I'm not sure why I would get an error via Windows but it works fine with the same account in Mgmt Studio.

imstac73

ASKER

I fixed the invalid name issue but now I am getting a time out expired error.

The stored procedure takes 1:12 minutes to run. I have set my connection timeout on my connection string to 1200. Is this not long enough?

imstac73

ASKER

When I try to preview the data in the dataset it gives me an additional error.

Null Value is eliminated by an aggregate or other SET operation.

Russell Fox

Try changing "COUNT(x.ICN)" to "COUNT(COALESCE(x.ICN, 0))". It's warning you that a grouping operation is eliminating some NULL values that you might actually want. It's just a warning, not an error, so you can also ignore it if you're getting the expected results.

imstac73

ASKER

I went ahead and just changed how I was pulling the data into the Windows application. I used the original stored procedure and made the actual table(s) with the data I needed in the application normal tables. Then in my Windows application I run the stored procedure using sql command and added my data table to the dataset and used the data from the table.

This actually increased the performance and fixed my error issue.