Archives of the TeradataForum

Message Posted: Wed, 30 Apr 2003 @ 20:54:38 GMT

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

<-- Anonymously Posted: Wednesday, April 30, 2003 14:56 -->

Hello All,

We're starting to run into some limitations using Mload (concurrent loads to the same table, etc) and we're wondering about what the general performance differences and limitations are between TPump and MLoad. At what number of rows does TPump drop out of contention under normal circumstances? I realize that it will be somewhat dependant on our environment, so I'm just looking for more general feedback on what everyone else is taking into consideration when deciding whether to use one tool or the other from number of rows, to number of updates vs inserts, to size of the table (I'm familiar with the index, trigger, etc limitations between the two - or "the rules" or the tools, so to speak).

We often load small amounts of data (50 to 100000 records into staging tables (no updates) and thought that those might be good TPump candidates. Some of the tables have as few as 5 columns, although some are large.

We will ultimately be using Informatica to execute the load operation - so we may not have some of the "configurability" that we might have in "true scripting" - if that impacts anything from your view. This also means that we don't care about which can summarize, aggregate, calculate, or access multiple sources because Informatica will be handling that portion of the transformation.

Thanks in advance for any insight on this.

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