 |
 |
Archives of the TeradataForum
Message Posted: Fri, 12 Sep 2003 @ 17:46:25 GMT
| Subj: | | Re: Will an Index be used with Derived Tables |
| |
| From: | | Walter, Todd A |
Sorry Michael but this statement is not correct. The column order in the derived table is irrelevant (unlike the rule for CREATE TABLE
where if no PI is specified TD defaults to the first column). The "extra" redistributions have been reduced by OPT features in various 4.x
releases and should be eliminated completely in 5.0 with the derived table optimization feature. The only DTs that should create a separate
spool in 5.0 are ones which force it with the operation inside the DT (eg aggregation and ordered analytics).
| |
