commit question

[admin]

Hi!
I use DataStage version 4.1 and UDB V7 tables (udb plugin/odbc stage)

> Can someone help me with the following problem:
>
I have 1 input table (read ordered by primary key), a simple transformation stage and 2 output tables. One table receives the input records and the other one is used as a restart table. I would like to commit on the first table after every n (10000) inserts/updates and at the same time (within
the same unit of work, commit) write the unique key of the last processed
row to the restart table. At the end of my job the restart_table should contain the last processed row (even when the number of rows is less than n
records).


Many thanks! Regards,
An