How to identify the erroneous rows in the source that are in ET table for a Mload table which is done through Informatica. Is it the exact location of the client row data i.e., say if source sequence in ET table is 570, then should I look into the 570 row of the source.
Say if the errlimit is set to 10 and there are only 8 rows in ET table and the job succeeds.
Will the source file or named pipe exists to identify the erroneous records or should we refer directly to the base table which is read by Informatica.
If it is from source table, how to identify the exact row that caused the issue using sourcesequence? (without the where clause with column value in ET) Should I sort the table according to the PI (or PPI and PI) and then find the 570th row using the row_num() or can it be directly pointed to the table without order by clause of 570. If so, how far is a select reliable to extract the 570th row directly from the table without order by clause?
Please refer to "Handling Teradata MultiLoad Errors" section in MultiLoad manual for further information, there is an example on how to determine the cause of errors: