Select multiple columns but group by only one - sql-server

Question

Using SQL Server 2008. I have been researching this problem for days. Thought about CTE and trying to use a predicate on the problem tables.
Nothing I have tried or researched has worked so here I am. The problem is it's returning duplicate OrderID.
I've marked the problem joins.
Even tried OUTER APPLY but it caused certain searches not to work. Tried INNER APPLY then duplicates again.
The problem joins, the tables, have multiple references to OrderID.
So tblRun has multiple rows with the same OrderID showing which run it was in at and what date and so forth.
I really need suggestions from all the guru's out there.
Here is the SQL:
DECLARE #CompanyID INT = 22718,
#StartDate DATETIME = '',
#EndDate DATETIME = '',
#SalesRepID INT = NULL,
#AssignedTo INT = NULL,
#ServiceDefID INT = NULL,
#ProductName VARCHAR(512) = NULL,
#IsCCOrder BIT = NULL,
#OrderID INT = NULL,
#LocationID INT = NULL,
#SalesRepLocationID INT = NULL,
#PONum VARCHAR(150) = NULL,
#InvoiceID INT = NULL,
#IsSplitOrder BIT = NULL,
#ContactID INT = NULL,
#ContactName VARCHAR(150) = NULL,
#JobName VARCHAR(200) = NULL,
#Notes VARCHAR(MAX) = NULL,
#CompanyName VARCHAR(255) = NULL,
#DueDateFrom DATETIME = '',
#DueDateTo DATETIME = '',
#SubmitedDateFrom DATETIME = '',
#SubmitedDateTo DATETIME = '',
#IsRush BIT = NULL,
#Msg VARCHAR(1000) = NULL
--#Stages dbo.int_tbltype READONLY
DECLARE #Stages TABLE (ID INT)
--INSERT INTO #Stages (ID) VALUES (1)
DECLARE #DueDate DATETIME = NULL
SET NOCOUNT ON
DECLARE #OrderIDsTBL TABLE(OrderID INT)
IF #Msg IS NOT NULL
BEGIN
INSERT INTO #OrderIDsTBL (OrderID)
SELECT OrderID
FROM tblOrderLog
WHERE Msg LIKE '%' + #Msg + '%'
END
IF #OrderID IS NOT NULL
BEGIN
INSERT INTO #OrderIDsTBL (OrderID)
VALUES (#OrderID)
END
DECLARE #OderIDsCnt INT = (SELECT COUNT(OrderID) FROM #OrderIDsTBL)
DECLARE #StageCnt INT = (SELECT COUNT(ID) FROM #Stages)
SELECT
o.OrderID,
o.CompanyID,
co.Name AS CompanyName,
o.ContactID,
o.JobName,
p.FirstName + ' ' + p.LastName AS ContactName,
p2.FirstName + ' ' + p2.LastName AS SalesRep,
o.DueDate,
CASE WHEN MAX(oi.PriorityService) > 0 THEN 1 ELSE 0 END AS IsRush,
ISNULL(s.StageID, 0) AS StageID,
o.Notes, r.SubmittedComplete,
dbo.fOrderRunLocationCSVByOrderID(o.OrderID) AS LocationCSV,
(SELECT
STUFF((SELECT DISTINCT ' ' + st.Name + '<br />' FROM tblStage st
INNER JOIN tblOrderItem oi ON oi.OrderID = o.OrderID
INNER JOIN tblRun r ON r.OrderItemID = oi.OrderItemID
INNER JOIN tblStage s ON s.StageID = r.StageID
LEFT JOIN tblRunService rs ON rs.RunID = r.RunID
WHERE (s.StageID = st.StageID)
AND (rs.AssignedTo = #AssignedTo OR #AssignedTo IS NULL)
FOR XML PATH(''), TYPE).value('.','VARCHAR(max)'), 1, 1, ''))
AS Stages,
Row_Number() Over(Order By o.OrderID Desc) As RowNum
FROM
tblOrder o
INNER JOIN
tblCompany co ON co.CompanyID = o.CompanyID
INNER JOIN
tblParty p ON p.PartyID = o.ContactID
-------- PROBLEM JOINS ------------
LEFT JOIN
tblOrderItem oi ON oi.OrderID = o.OrderID
LEFT JOIN
tblRun r ON r.OrderItemID = oi.OrderItemID
LEFT JOIN
tblService srv ON srv.OrderItemID = oi.OrderItemID
-------- END PROBLEM JOINS ------------
LEFT JOIN
tblStage s ON s.StageID = r.StageID
LEFT JOIN
tblParty p2 ON p2.PartyID = o.SalesRepID
LEFT JOIN
tblEmployee e ON e.EmployeeID = o.SalesRepID
LEFT JOIN
tblShipTo st ON o.ShipToID = st.ShipToID
WHERE
(#CompanyID IS NULL OR (o.CompanyID = #CompanyID )) AND
(#IsCCOrder IS NULL OR (ISNULL(o.IsCreditCardOrder, 0) = #IsCCOrder )) AND
(#SalesRepID IS NULL OR o.SalesRepID = #SalesRepID) AND
(#ServiceDefID IS NULL OR (srv.ServiceDefID = #ServiceDefID)) AND
(#ProductName IS NULL OR (oi.Name LIKE '%' + #ProductName + '%')) AND
(#IsSplitOrder IS NULL OR (#IsSplitOrder = 1 AND oi.SplitOrderID IS NOT NULL)) AND
(
(#StartDate = '' OR #EndDate = '') OR
(#StartDate >= CreatedDate AND #EndDate <= COALESCE(CancelledDate, ClosedDate, GetDate())) OR
(#StartDate <= COALESCE(CancelledDate, ClosedDate, GETDATE()) AND #EndDate >= COALESCE(CancelledDate, ClosedDate, GetDate()) ) OR
(#StartDate <= CreatedDate AND #EndDate >= CreatedDate )
) AND
(#LocationID IS NULL OR (#LocationID = srv.LocationID OR srv.LocationID IS NULL)) AND
(#SalesRepLocationID IS NULL OR (#SalesRepLocationID = e.LocationID OR e.LocationID IS NULL))
AND (#InvoiceID IS NULL OR o.InvoiceID = #InvoiceID )
AND (#PONum IS NULL OR o.PONum LIKE '%' + #PONum + '%')
AND (COALESCE(s.StageID, 0) IN (SELECT ID FROM #Stages) OR #StageCnt = 0)
AND (o.ContactID = #ContactID OR #ContactID IS NULL)
AND (p.FirstName + ' ' + p.LastName LIKE '%' + #ContactName + '%' OR #ContactName IS NULL)
AND (o.JobName LIKE '%' + #JobName + '%' OR #JobName IS NULL)
AND (o.Notes LIKE '%' + #Notes + '%' OR #Notes IS NULL)
AND (co.Name LIKE '%' + #CompanyName + '%' OR #CompanyName IS NULL)
AND (o.DueDate >= #DueDateFrom OR #DueDateFrom = '')
AND (o.DueDate <= #DueDateTo OR #DueDateTo = '')
AND (r.SubmittedComplete >= #SubmitedDateFrom OR #SubmitedDateFrom = '')
AND (r.SubmittedComplete <= #SubmitedDateTo OR #SubmitedDateTo = '')
AND (#IsRush = (CASE WHEN oi.PriorityService > 0 THEN 1 ELSE 0 END)
OR #IsRush IS NULL)
AND (o.OrderID IN (SELECT OrderID FROM #OrderIDsTBL) OR #OderIDsCnt = 0)
GROUP BY
o.OrderID, o.CompanyID,
co.Name,
o.ContactID, o.JobName,
p.FirstName, p.LastName, p2.FirstName, p2.LastName,
o.DueDate, o.Notes,
r.SubmittedComplete,
s.StageID
Thanks for any suggestions. I've been working on this for some time now and just can't get it working right.

Answer 1

It looks like you're trying to do too much with a single SELECT statement. If you want one row per unique OrderID, then don't join with the tables that have multiple rows for the same OrderID. Remove the GROUP BY clause. Use one or more separate SELECT statements to get the details from the tables that have multiple rows per OrderID.

Answer 2

Thanks everyone for the suggestions but I found my own solution.
With the entire sql, I placed it into a temp table.
Then used this to sort out the duplicate OrderIds....
SELECT OrderID, CompanyID, ContactID, CompanyName, JobName, ContactName,
SalesRep, DueDate, IsRush , StageID, Notes, SubmittedComplete,
LocationCSV, Stages
FROM (SELECT *, ROW_NUMBER() OVER(PARTITION BY OrderID ORDER BY OrderID DESC) 'RowRank'
FROM #SearchTbl
)sub
WHERE RowRank = 1