Please see if constraint-based load ordering property was enabled for the session. If you configure the session to use constraint-based loading, you must specify pass-through partitioning for all transformations that are downstream from the last active source.
Its a simple pass through mapping where we are generating the files(Data is huge). No constraint based loading. Overall record count is matching between DB and Informatica. But when we are checking record count for each partition wise it is not matching. That is what I am not able to understand?
How are you determining the inconstancy?
REP_SESS_TBL_LOG contains the numbers of records written to each target instance per session partition. Multiply this number by the average size of each record in bytes, then you get an estimate for the amounts of data written.
This is the best that you can get without parsing the session log files. The session log file is the only place where the exact numbers are saved; this information is not available anywhere in the repository, and the numbers indicated by the Integration Service (e.g. in the Workflow Monitor) are not kept anywhere.
REP_SESS_TBL_LOG view contains information on the status of an individual session run against a target. It provides the last update time, row counts and error status based on the last update timestamp on per target basis. Let us consider a session with one target. The target has 12 partitions. After the session run, the view should have one row for each target partition, but the view has an inconsistently large number of rows, thereby deviating from the expected behavior.
Please find below the records count partition wise,
DB Count Infa Count
Total 6976 6976
Any thought on it?
Thanks & Regards,