 |
 |
Archives of the TeradataForum
Message Posted: Mon, 07 Oct 2002 @ 12:30:07 GMT
| Subj: | | Re: Fastexport-Outmod and RDBMS-Restarts |
| |
| From: | | Geoffrey Rommel |
| | We now encountered some network problems, the connection to the database broke down and fexp re-sent the SQL statement to the
database (with the message "UTY8720 Re-starting this FastExport task because the RDBMS crashed"), but the exported data was corrupt because
the Outmod did not get a "restart" signal.... | |
| | Is there any way to avoid this? How can I recognize such a situation and react on it when the "correct" entry code is not sent (or
did I miss something important)? | |
Jawohl! When FastExport is restarting like this, it will pass an Entry Code of 5 to the OUTMOD. See the FastExport Reference (B035-
2410-061A, June 2001), p. 2-39. Your code should look at the Entry Code and decide what to do. Of course, one thing you could do is simply
print a message saying "Sorry, I can't restart -- resubmit the job from the beginning" and exit (that's what I do). Or you could take
checkpoints every so often, write them to a file, and then restart from the last checkpoint. It's up to you.
--wgr
| |
