December 12, 2019 at 6:19 pm
Hi,
Am using merge sp to insert or update records in target table.Below is the sp
MERGE Products AS TARGET
USING UpdatedProducts AS SOURCE
ON (TARGET.ProductName= SOURCE.ProductName)
--When records are matched, update the records if there is any change
WHEN MATCHED
THEN UPDATE SET TARGET.Rate = SOURCE.Rate
--When no records are matched, insert the incoming records from source table to target table
WHEN NOT MATCHED BY TARGET
THEN INSERT (ProductID, ProductName, Rate) VALUES (SOURCE.ProductID, SOURCE.ProductName, SOURCE.Rate)
In my source table we have two records
Source table :
ProductID ProductName Rate
101 TEA 10.00
102 TEA 25.00
Target table: empty table
ProductID ProductName Rate
try to move source records to target with merge sp. but two records are inserted.
ProductID ProductName Rate
101 TEA 10.00
102 TEA 25.00
It should be one record in target table. because first record will match with second record.
Expected result.
ProductID ProductName Rate
101 TEA 25.00
Can you please help,How merge statement pass records to condition(ON (TARGET.ProductName= SOURCE.ProductName) one by one or bulk?If one by one, Why second record also inserted in target table.
December 12, 2019 at 6:41 pm
I would guess that it's because ProductName is not unique in the table
Try changing your match criteria from this
(TARGET.ProductName= SOURCE.ProductName)
to this
(TARGET.ProductId= SOURCE.ProductId)
The absence of evidence is not evidence of absence
- Martin Rees
The absence of consumable DDL, sample data and desired results is, however, evidence of the absence of my response
- Phil Parkin
December 12, 2019 at 7:07 pm
You will find you get much better performance if you split the merge into two separate statements (an update followed by an insert).
Performance Tip: The conditional behavior described for the MERGE statement works best when the two tables have a complex mixture of matching characteristics. For example, inserting a row if it doesn't exist, or updating a row if it matches. When simply updating one table based on the rows of another table, improve the performance and scalability with basic INSERT, UPDATE, and DELETE statements. For example:
INSERT tbl_A (col, col2)
SELECT col, col2
FROM tbl_B
WHERE NOT EXISTS (SELECT col FROM tbl_A A2 WHERE A2.col = tbl_B.col);
https://docs.microsoft.com/en-us/sql/t-sql/statements/merge-transact-sql?view=sql-server-ver15
December 15, 2019 at 4:49 pm
I tried with (TARGET.ProductId= SOURCE.ProductId) . But same behavior two records are inserted. if we re-execute the sp. that time it will updating. Please hep on this if any other way to pass one by one records.
December 15, 2019 at 6:35 pm
Is it really necessary to use a merge statement? Why not something like this?
DROP TABLE IF EXISTS Test_Products;
DROP TABLE IF EXISTS Test_UpdatedProducts;
GO
CREATE TABLE Test_Products(ProductIDint, ProductName varchar(10), Rate int);
CREATE TABLE Test_UpdatedProducts(ProductIDint, ProductName varchar(10), Rate int);
GO
INSERT Test_UpdatedProducts VALUES
(101, 'TEA', 10.00),
(102, 'TEA', 25.00);
GO
WITH UP_CTE AS(
SELECT
ProductID, ProductName, Rate,
ROW_NUMBER() OVER (PARTITION BY ProductName ORDER BY ProductID DESC) AS RowNum
FROM Test_UpdatedProducts UP)
INSERT Test_Products(ProductID, ProductName, Rate)
SELECT
ProductID, ProductName, Rate
FROM UP_CTE
WHERE
RowNum=1
EXCEPT
SELECT * FROM Test_Products;
SELECT * FROM Test_Products;
Aus dem Paradies, das Cantor uns geschaffen, soll uns niemand vertreiben können
December 15, 2019 at 8:13 pm
it is working as expected - you are matching on product name - "TEA" - no record exists so both are inserted.
Note that the error would happen even with a straight update as source would have 2 rows to update a single destination row so regardless of the method used you need to change how you process the source.
sample to reproduce the error with merge
if object_id('tempdb..#Products ') is not null
drop table #Products;
create table #Products
( ProductID int
, productname varchar(10)
, rate decimal(10, 2)
)
;
if object_id('tempdb..#UpdatedProducts ') is not null
drop table #UpdatedProducts;
create table #UpdatedProducts
( ProductID int
, productname varchar(10)
, rate decimal(10, 2)
)
;
insert into #UpdatedProducts
select 101
, 'tea'
, 10;
insert into #UpdatedProducts
select 102
, 'tea'
, 25;
merge #Products as target using #UpdatedProducts as source
on (target.productname = source.productname)
--When records are matched, update the records if there is any change
when matched
then update
set target.rate = source.rate
--When no records are matched, insert the incoming records from source table to target table
when not matched by target
then insert
(ProductID
, productname
, rate
)
values (source.ProductID, source.productname, source.rate
);
/*
do same merge again to prove that doing a merge like this, where there are 2 records on the source and we try and update the "same" record on the target
*/
merge #Products as target using #UpdatedProducts as source
on (target.productname = source.productname)
--When records are matched, update the records if there is any change
when matched
then update
set target.rate = source.rate
--When no records are matched, insert the incoming records from source table to target table
when not matched by target
then insert
(ProductID
, productname
, rate
)
values (source.ProductID, source.productname, source.rate
);
go
/*
Above fail with error as expected as it is trying to update the same record twice
Msg 8672, Level 16, State 1, Line 55
The MERGE statement attempted to UPDATE or DELETE the same row more than once. This happens when a target row matches more than one source row. A MERGE statement cannot UPDATE/DELETE the same row of the target table multiple times. Refine the ON clause to ensure a target row matches at most one source row, or use the GROUP BY clause to group the source rows.
*/
/*
if only one record for a single productname is supposed to exist on the target table then we need to filter the source
how that is filtered depends on business rules - for this example I assumed that the higher product id is to be used
*/
truncate table #Products
merge #Products as target using
(
select t.ProductID
, t.productname
, t.rate
from (
select ProductID
, productname
, rate
, row_number() over (partition by productname
order by ProductID desc
) rownum
from #UpdatedProducts
) t
where t.rownum = 1
) as source
on (target.productname = source.productname)
--When records are matched, update the records if there is any change
when matched
then update
set target.rate = source.rate
--When no records are matched, insert the incoming records from source table to target table
when not matched by target
then insert
(ProductID
, productname
, rate
)
values (source.ProductID, source.productname, source.rate
);
select *
from #Products
-- insert new record so we know update worked
insert into #UpdatedProducts
select 103
, 'tea'
, 225;
merge #Products as target using
(
select t.ProductID
, t.productname
, t.rate
from (
select ProductID
, productname
, rate
, row_number() over (partition by productname
order by ProductID desc
) rownum
from #UpdatedProducts
) t
where t.rownum = 1
) as source
on (target.productname = source.productname)
--When records are matched, update the records if there is any change
when matched
then update
set target.rate = source.rate
--When no records are matched, insert the incoming records from source table to target table
when not matched by target
then insert
(ProductID
, productname
, rate
)
values (source.ProductID, source.productname, source.rate
);
select *
from #Products December 15, 2019 at 8:54 pm
For your example you would be much better off, in terms of performance, by having an update followed by an insert:
UPDATE p
SET p.rate = up.rate
FROM #Products p
INNER JOIN #UpdatedProducts up
ON up.productname = p.productname
INSERT INTO #Products(ProductID, productname, rate)
SELECT up.ProductID,
up.productname,
up.rate
FROM #UpdatedProducts up
WHERE NOT EXISTS(SELECT * FROM #Products p WHERE p.productname = up.productname)
December 15, 2019 at 9:59 pm
For your example you would be much better off, in terms of performance, by having an update followed by an insert:
UPDATE p
SET p.rate = up.rate
FROM #Products p
INNER JOIN #UpdatedProducts up
ON up.productname = p.productname
INSERT INTO #Products(ProductID, productname, rate)
SELECT up.ProductID,
up.productname,
up.rate
FROM #UpdatedProducts up
WHERE NOT EXISTS(SELECT * FROM #Products p WHERE p.productname = up.productname)
I wholeheartedly agree with that. I'll also suggest that since they're changing "rates" on products, they really should look into Type 6 SCDs, as well. Temporal Tables make some pretty short work of that.
--Jeff Moden
Change is inevitable... Change for the better is not.
December 15, 2019 at 10:58 pm
Jonathan AC Roberts wrote:For your example you would be much better off, in terms of performance, by having an update followed by an insert:
UPDATE p
SET p.rate = up.rate
FROM #Products p
INNER JOIN #UpdatedProducts up
ON up.productname = p.productname
INSERT INTO #Products(ProductID, productname, rate)
SELECT up.ProductID,
up.productname,
up.rate
FROM #UpdatedProducts up
WHERE NOT EXISTS(SELECT * FROM #Products p WHERE p.productname = up.productname)I wholeheartedly agree with that. I'll also suggest that since they're changing "rates" on products, they really should look into Type 6 SCDs, as well. Temporal Tables make some pretty short work of that.
update/insert does not error - but insert still adds the same 2 rows - and the update will set them both to the same value - not necessarily the correct one so the base logic error still remains.
December 15, 2019 at 11:36 pm
Jeff Moden wrote:Jonathan AC Roberts wrote:For your example you would be much better off, in terms of performance, by having an update followed by an insert:
UPDATE p
SET p.rate = up.rate
FROM #Products p
INNER JOIN #UpdatedProducts up
ON up.productname = p.productname
INSERT INTO #Products(ProductID, productname, rate)
SELECT up.ProductID,
up.productname,
up.rate
FROM #UpdatedProducts up
WHERE NOT EXISTS(SELECT * FROM #Products p WHERE p.productname = up.productname)I wholeheartedly agree with that. I'll also suggest that since they're changing "rates" on products, they really should look into Type 6 SCDs, as well. Temporal Tables make some pretty short work of that.
update/insert does not error - but insert still adds the same 2 rows - and the update will set them both to the same value - not necessarily the correct one so the base logic error still remains.
I don't think the OP has explained very precisely what the problem is. Why do they only want ProductId 101? Why does this take presidency over ProductId 102? Why does the OP want 20.0 not 10.0 in the rate?
December 15, 2019 at 11:45 pm
I don't think the OP has explained very precisely what the problem is. Why do they only want ProductId 101? Why does this take presidency over ProductId 102? Why does the OP want 20.0 not 10.0 in the rate?
Ah yes the OP wanted the ProductId 101 with the 102 Rate. My code was not at all correct. Clarification would help. It all seems backwards really, why is there an update table and why is the Products table not updated directly?
Aus dem Paradies, das Cantor uns geschaffen, soll uns niemand vertreiben können
Viewing 11 posts - 1 through 10 (of 10 total)
You must be logged in to reply to this topic. Login to reply