Re: help! cannot understand "leading subset of columns" in index usage

let's say there is an index on (c1,c2), and the index is 50 times smaller than the table. Just think: will it be faster to scan the index to satifsfy
where c2=10 clause, rather then to scan the whole table? Even forgetting about skip scans for a while, just scanning a 50 times smaller index vs. scanning the whole wide table?

What do you think? Received on Mon Jun 20 2005 - 15:05:31 CDT