Archives of the TeradataForum
Message Posted: Thu, 11 Jan 2007 @ 16:08:54 GMT
| Subj: || || Re: DataStage Enterprise Edition |
| From: || || Geoffrey Rommel |
| ||2) Make sure you set the load utilities to run serially and not in parallel, all the stages prior to the load can run in parallel, but
don't try to run the load utilities in parallel.|| |
To expand a little on this point, you need to pay close attention to the parallel option (Hash, for instance) used by each stage. Test the job
in non-parallel mode (i.e. with one thread) and then in parallel mode to make sure you get the same results. As Robert says, the actual load stage
doesn't need to be parallel.
But why am I telling you this? Isn't there an IBM consultant helping you with this development? If not, there should be.