Archives of the TeradataForum
Message Posted: Fri, 14 Mar 2008 @ 13:26:14 GMT
| Subj: || || Re: Join with column faster than join with primary index column? |
| From: || || Anomy Anom |
<-- Anonymously Posted: Fri, 14 Mar 2008 06:58 -->
You only included one step out of the query plan, so it's hard to answer your question as to why the total time is longer.
For the one query step you included, the join to Spool 4 is faster because only required columns (and rows) are projected into spool so
scanning Spool 5 reads fewer blocks than scanning the original table. One would expect that improvement would be at least partially offset by the
cost of building Spool 5 in the first place.