Execute SQL Task Error Handling

I am using the Execute SQL Task to call a sproc which I am not in the design stream for, so I can call it but I cannot change it. I need to deal with the issue of handling an error from a call to this sproc- the stored procedure does not return any output variables nor a return value that I can nail, all I get is a failure in the task. The sproc itself is just a straightforward insert into a table which can fail for any of the obvious reasons (PK violation, bad data, null data et al). The point is I want to know which rows failed in the insert (moving data from one databases table to another one row at a time wiith the sproc being inside a loop task over a recordset variable populated via a data flow call to a SELECT statement) so that I can flag then as such so that when I am finished in my loop I am left with a bunch of rows that did not make the trip from database A to databases B - within the loop, for each row that the sproc call is successfull, will hjave another sproc follow it to delete it from the original, but I cannot get this far when my INSERT sproc call fails. I have tried changing the MaxErrorCount to a big number - ceratinly more than I need but it does not seem to matter .lCan you help me out here - or is the design flawed anyway?