Hello Support,
We use your OLEDB provider to synchronise data between our SQL Server OLTP database and a Redshift data warehouse.
In our data flow there are few occurrences where we use UPDATE in an OLEDB command in a data flow.
In many cases when DATE or DATETIME columns are updated, resulting values in the destination are completely different from what's specified in the query. For instance, we run the following code (caught by PGNP profiler):
Code:
---------- Client SQL 2016/02/01 15:29:30.230 D27BC981 ------------
UPDATE inventorytransfer
SET
fromstoreid = ?,
tostoreid = ?,
status = ?,
createdate = ?,
transferedby = ?,
receivedby = ?,
cancelledby = ?,
transferdate = ?,
receivedate = ?,
canceldate = ?,
"__$etl_lsn" = ?
WHERE inventorytransferid = ?;
---------- Executed SQL Parse:0.000 Prepare:0.000 Execute:1307.024 GetRows:0.000 Rows:0 ---------
EXECUTE "ae7c76d7f34ea77c5b728157b2de8bd1"([DBTYPE_I8,613],[DBTYPE_I8,401],[DBTYPE_WSTR,"Cancelled"],[DBTYPE_DBTIMESTAMP,2015-09-10 15:38:09.447],[DBTYPE_WSTR,"Richie"],[DBTYPE_WSTR,"Joshua"],[DBTYPE_WSTR,"RDS Test"],[DBTYPE_DBTIMESTAMP,2016-02-01 15:26:27.360],[DBTYPE_DBTIMESTAMP,2015-09-10 15:43:41.533],[DBTYPE_DBTIMESTAMP,2016-02-06 11:56:57.000],[DBTYPE_I8,245680],[DBTYPE_WSTR,"0000087B000368760024"])
But when we check the destination on completion, we find following DATETIME values there:
createdate = 2015-09-25 11:16:20
transferdate = 2016-02-16 11:16:20
receivedate = 2015-09-25 11:16:20
canceldate = 2016-02-21 07:16:20
These are completely different dates, not present in the profiler trace.
Could you please propose some workaround for it and if it's a bug, expected date when it's going to be fixed?
Thank you very much in advance.
Best Regards,
Alexander.
