MERGE in SQL Server

Techies--

Will someone explain to me why this MERGE statement is setting the RecordIsDeleted flag to 1 on every row? Obviously NOT MATCHED by SOURCE trips it-- but I thought the criteria for the MATCH is on the 2 keys under the ON condition--not every column after the ON, which is what it appears to be doing.  Any advice on how to correct this?



MERGE dimension.CustomerSegment  AS TARGET

	-- SELECT ALL COLUMNS IN THE SOURCE ETL TABLE
	USING ( SELECT DS.SegmentKey
				 , DC.CustomerKey
				 , DCS.[MemberInternalKey]
				 , DCS.[BuyingUnitInternalKey]
				 , DCS.[SegmentInternalKey]
				 , DCS.[MemberDK]
				 , DCS.[BuyingUnitDK]
				 , DCS.[SegmentId]
				 , DCS.[SegmentDescription]
				 , DCS.[MatrixMemberId]
				 , DCS.[ClubInternalKey]
				 , DCS.[UpdatedDate]
				 , DCS.[StartDate]
				 , DCS.[EndDate]
				 , DCS.[AttachmentSourceId]
				 , DCS.[AttachmentSourceName] 
				 , DCS.[RecordFromDate]
				 , DCS.[RecordToDate]
				 , DCS.[RecordIsCurrent]
				 , DCS.[RecordisDeleted]        
			  FROM [etl].[CustomerSegment]		DCS
			 INNER JOIN dimension.Customer		 DC		ON DC.MemberDK				=	DCS.MemberDK
													   AND DC.BuyingUnitDK			=	DCS.BuyingUnitDK
			 INNER JOIN dimension.Segment		 DS		ON DS.SegmentInternalKey	=	DCS.SegmentInternalKey

          ) AS Source
                 ( SegmentKey
				 , CustomerKey
				 , [MemberInternalKey]
				 , [BuyingUnitInternalKey]
				 , [SegmentInternalKey]
				 , [MemberDK]
				 , [BuyingUnitDK]
				 , [SegmentId]
				 , [SegmentDescription]
				 , [MatrixMemberId]
				 , [ClubInternalKey]
				 , [UpdatedDate]
				 , [StartDate]
				 , [EndDate]
				 , [AttachmentSourceId]
				 , [AttachmentSourceName]
				 , [RecordFromDate]
				 , [RecordToDate]
				 , [RecordIsCurrent]
				 , [RecordisDeleted]
				 )
		ON (    Source.CustomerKey		= Target.CustomerKey		-- Key will be customer, segment, and segment start date.	
		         AND Source.SegmentKey	= Target.SegmentKey			-- Remember that segments can come and go
			
	

		   )

		-- NO MATCH IN TARGET MEANS NEW DATA - PERFORM AN INSERT
		WHEN NOT MATCHED BY TARGET THEN
	    INSERT (   [SegmentKey]
				  ,[CustomerKey]
				  ,[UpdatedDate]
				  ,[StartDate]
				  ,[EndDate]
				  ,[AttachmentSourceId]
				  ,[AttachmentSource]
				  ,[RecordFromDate]
				  ,[RecordToDate]
				  ,[RecordIsCurrent]
				  ,[RecordisDeleted]
				  ,[InsertedBy]
				  ,[InsertedDate]
				  ,[ModifiedBy]
				  ,[ModifiedDate]
		      )
        VALUES
             (   Source.[SegmentKey]
				,Source.[CustomerKey]
				,Source.[UpdatedDate]
				,Source.[StartDate]
				,Source.[EndDate]
				,Source.[AttachmentSourceId]
				,Source.[AttachmentSourceName]
				,Source.[RecordFromDate]
				,Source.[RecordToDate]
				,Source.[RecordIsCurrent]
				,Source.[RecordisDeleted]
			    ,@ExecutedBy						-- InsertedBy
			    ,GetDate()						    -- InsertedDate
			    ,@ExecutedBy						-- ModifiedBy
			    ,GetDate()						    -- ModifiedDate
		     )
		-- NO MATCH IN SOURCE MEANS WHAT SEGMENT DID EXIST FOR MEMBER IS NO LONGER CURRENT
		-- *Note: Because deleting/fragmenting indexes can be a performance killer, the row wil
		--        be flagged as RecordIsDeleted to TRUE.

		WHEN NOT MATCHED BY SOURCE THEN
		 UPDATE	
		   SET RecordIsDeleted		=	1   -- Set for future delete/reindex maint job.

		WHEN MATCHED			
         THEN UPDATE	
			SET Updateddate				=	Source.[Updateddate]		
			,StartDate					=	Source.[StartDate]			
			,EndDate					=	Source.[EndDate]
			,AttachmentSourceId			=	Source.[AttachmentSourceId]
			,AttachmentSource			=	Source.[AttachmentSourceName]
			,RecordToDate				=	Source.[RecordToDate]
			,[RecordIsCurrent]			=	Source.[RecordIsCurrent]
			,[RecordIsDeleted]			=	Source.[RecordIsDeleted]  
			,ModifiedBy					=	@ExecutedBy						-- ModifiedBy
			,ModifiedDate				=	GetDate()						--  ModifiedDate	


 ; -- END OF THE MERGE STATEMENT

Open in new window

Paula DiTalloIntegration developerAsked:
Who is Participating?

Improve company productivity with a Business Account.Sign Up

x
 
Kent OlsenConnect With a Mentor Data Warehouse Architect / DBACommented:
Hi Ditallop,

It looks like a data issue to me.  When you inner join the 3 tables to produce [source] are you sure that you're getting the rows that you expect?

I created a small test case that closely resembles your example.  It works just fine.


Kent


CREATE TABLE dbo.t1 (k1 integer, k2 integer, RecordIsDeleted integer, Text varchar (100));
CREATE TABLE dbo.t2 (k1 integer, k2 integer, Text varchar (100));

DELETE FROM dbo.t1;
DELETE FROM dbo.t2;

INSERT INTO dbo.t1 SELECT 1, 101, 0, 'A';
INSERT INTO dbo.t1 SELECT 2, 102, 0, 'B';
INSERT INTO dbo.t1 SELECT 3, 103, 0, 'C';
INSERT INTO dbo.t1 SELECT 4, 104, 0, 'D';
INSERT INTO dbo.t2 SELECT 1, 201, 'E';
INSERT INTO dbo.t2 SELECT 2, 102, 'F';
INSERT INTO dbo.t2 SELECT 5, 103, 'G';
INSERT INTO dbo.t2 SELECT 6, 204, 'H';

select * from t1;
select * from t2;

MERGE dbo.t1 AS [target]
USING 
( 
  SELECT k1, k2, Text
  FROM t2
) as [source] (k1, k2, text)
  ON ([source].k1 = [target].k1
 AND [source].k2 = [target].k2)
WHEN NOT MATCHED BY TARGET THEN
  INSERT (k1, k2, RecordIsDeleted, Text) 
    VALUES ([source].k1, [source].k2, 0, [source].text)
WHEN NOT MATCHED BY SOURCE THEN
 UPDATE	SET RecordIsDeleted		=	1   -- Set for future delete/reindex maint job.
WHEN MATCHED			
  THEN UPDATE	
    SET Text = [source].Text;
                        
                        
select * from t1;
select * from t2;

Open in new window


The results are:

t1 before the merge:
1      101      0      A
2      102      0      B
3      103      0      C
4      104      0      D

t2 before the merge:
1      201      E
2      102      F
5      103      G
6      204      H

t1 after the merge:
1      101      1      A
2      102      0      F
3      103      1      C
4      104      1      D
5      103      0      G
6      204      0      H
1      201      0      E

t2 after the merge:
1      201      E
2      102      F
5      103      G
6      204      H
0
 
Paula DiTalloIntegration developerAuthor Commented:
@hnasr, there isn't an odbc component to this particular issue.
0
 
hnasrCommented:
Unable to reproduce the issue.

Can you help me in reproducing the tables, so to understand and test the problem
0
What Kind of Coding Program is Right for You?

There are many ways to learn to code these days. From coding bootcamps like Flatiron School to online courses to totally free beginner resources. The best way to learn to code depends on many factors, but the most important one is you. See what course is best for you.

 
Paula DiTalloIntegration developerAuthor Commented:
Maddening isn't it? This is a perfectly good merge statement that should work--but it just plain doesn't with the live data--I've had to give up on this approach and go the update/insert/delete route... due to time constraints.  Thanks for looking at this.
0
 
Kent OlsenData Warehouse Architect / DBACommented:
My hunch still revolves around a data issue.  I'll be glad to keep working with you if you want to dig some more.
0
 
Paula DiTalloIntegration developerAuthor Commented:
Thanks Kdo--now that the immediate need to solve this has past, I'll resurrect the what I have and post a test script
0
Question has a verified solution.

Are you are experiencing a similar issue? Get a personalized answer when you ask a related question.

Have a better answer? Share it in a comment.

All Courses

From novice to tech pro — start learning today.