Column Collation Not Replicated with Parallel Load
Hello,
On Microsoft Sql Server, our source database collation is SQL_Latin1_General_CP1_CI_AS however we have column collations that are SQL_Latin1_General_CP1_CS_AS. If we define the replicate task without any transformations, it maintains this column collation. However if we add transformations, the column collation changes to SQL_Latin1_General_CP1_CI_AS.
I can understand this limitation when the table is manipulated, either transformations to existing columns or adding columns that subsequently change the table definition. Unfortunately this issue also occurs when we select a table for parallel load. As we have some larger source tables requiring the case sensitive collation, it would help our full load if we could utilize the parallel load.
We submitted a case support asking about this issue and it was determined this limitation was by design, however this wasn't clearly indicated in the documentation (p. 284 of the handbook https://help.qlik.com/en-US/replicate/November2020/pdf/Replicate%20Setup%20and%20User%20Guide.pdf says "When replicating from one Microsoft SQL Server database to another, column and table collations will be replicated to the target. ")
We have analyzed this further and determined that this is indeed a bug. We have reopened your support case 02088020 to handle this. Thanks for your patience!
