Hi all,
As part of a data driven dataload process, I'm providing users with a facility to look at errors that have occurred at a row level in the data cleaning process. i.e. those rows that failed the cleaning process and were left in the staging database, but not copied to the clean database.
Each table in the staging database has an additional column, DL_LoadErrorMessage which indicates the reason why the row failed cleaning, and is empty otherwise.
Part of my dataload reference data are tables:
DL_System - indicates details about the source system, the staging and clean databases
DL_Table - a row per table. Is a child table of DL_System. Contains rules IN SQL snippets.
DL_Field - a row per field. Is a child table of DL_Table. Contains rules and field validation in SQL snippets.
The dataload processes uses these three tables to perform the entire dataload in an entirely data driven way using dynamic SQL.
I have a series of reports that then let the user drill down from system level to look at what staging errors occurred and how many rows failed for which reason. I then want a click through to the rows that errors for that table. I just want to wrote on report that will show all the rows on a staged table where DL_LoadErrorMessage is equal to a specific value. The SQL that returns this from the server will be constructed in dynamic SQL and is of the form 'SELECT * FROM '+ @TableName + ' WHERE DL_LoadErrorMessage = ''' + @ErrorMessage +''''
This will give the user the ability to then go back to the source system and address the error (mostly Informix dates, currently).
I have no idea what any of the column names might be at report design time, their type, or the number of them. The only thing I do know is that they will all have a DL_LoadErrorMessage column.
Is there any way in SSRS where I can supply a dataset to a control that may have an entirely different structure each time it is called and it will derive the column names and populate itself accordingly?
Yours, in hope,
Mark
PS. SSRS 2008R2.