Archives of the TeradataForum

Message Posted: Mon, 18 Mar 2002 @ 14:06:11 GMT

	<Prev	Next>		<<First	<Prev	Next>	Last>>

Karen,

We had similar discussions here for one of our source areas. We actually implemented the approach you described in that we moved data to the mainframe (MVS) and then loaded to Teradata using a common strategy. In theory it was a good idea but the reality was we introduced more potential points of failure into the etl process. We still had to support multiple ways of moving data to the mainframe. We added steps to the etl process in that the same data was moved twice. once to the mainframe and then once to Teradata.

We rewrote the process using a shortest path methodology and have been much happier. We added a NT server to support the SQL server environment and run informatica, and mload from it using Maesto as the scheduler.

Jim

	<Prev	Next>		<<First	<Prev	Next>	Last>>