|
Archives of the TeradataForumMessage Posted: Thu, 23 Feb 2006 @ 19:23:38 GMT
Unless you use large 'pack' values there probably wont be a lot of performance gain using param arrays over multi-statement Import queries. One advantage of param arrays is that you are limited to how many statements you can specify in a multi-statement query, due to the limit of 2000 parameters in a query. Someone just gave me performance tests showing that even if you use single statements you can acheive maybe 150 rows/sec with 100 bytes per row. If you increase the number of statements in the query to 20 you may get 2000/sec, so even fairly large Imports should be quite fast. There has not been a request for parameter arrays in SQLA. I was not aware that customers used SQLA to load batches of 250,000 rows, so I had not considered this to be a requirement. (Once you get up to that kind of size BTEQ would definately provide a better performance since you can use multiple sessions in addition to using parameter arrays.) Mike Dempsey
| ||||||||||||||||||||||||||||||||||||||||||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||
Copyright 2016 - All Rights Reserved | ||||||||||||||||||||||||||||||||||||||||||||||||
Last Modified: 15 Jun 2023 | ||||||||||||||||||||||||||||||||||||||||||||||||