|
Archives of the TeradataForumMessage Posted: Wed, 14 Oct 2009 @ 10:11:47 GMT
Boris Mogilevsky wrote:
- You can store a larger range of values within the same number of bytes in an integer - You might specify an exact number of digits for a decimal - Calculations on integers are truncated, whereas decimals are rounded
see above. For PI there's no difference because integer and (dec(n,0)) hash to the same value.
Maybe this is from V2R1 when there was no BIGINT yet. An integer might easily run out of number when the default batch value of 100.000 is used: 2000 AMPs * 100.000 = 200.000.000 After 10 restarts it hits the limit. So the actual decision is just BIGINT vs. DEC(18,0). Which one to choose, see above. Dieter
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright 2016 - All Rights Reserved | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Last Modified: 15 Jun 2023 | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||