Archives of the TeradataForum
Message Posted: Thu, 20 Feb 2003 @ 14:19:22 GMT
I've never come across the regular use of RM before, I've usually only used that when I know (or expect) that a previous job has failed - - usually in a development environment.
If you still think you have a need to run this step, then certainly checking for the error/work/log tables and only running the RM if they're present could be a good mechanism for you.
But of course that assumes that the tables haven't been cleared up and that they are named accoring to whatever convention you've coded into your check script...
You may also be able to use the LOCK NOWAIT option with an ACCESS lock (I think that was available in v2r3) and simply trap the return from that.
If it fails with a "I can't get the requested lock" then that must be because the existing lock is EXCLUSIVE, you then have to assume that's because the mload locks are still in place and hence run your RM command. This doesn't require that the work/error tables are left in place and it doesn't rely on the mload code meeting certain standards (or the standards never changing).
|Copyright 2016 - All Rights Reserved|
|Last Modified: 27 Dec 2016|