1 Things that could be done to DBMS -- 12 October 1998

     2 

     3 Query optimizer:

     4 

     5 	Optimize for clustering indexes (use S/T and G/H plans)

     6 		e.g. assume Auto-increment column => clustering

     7 		or measure the clustering in the stats

     8 

     9 	Optimize for multi-column lookup

    10 		e.g. For "a=v1 and b=v2" with index on (a,b)

    11 

    12 	Optimize for like-predicates

    13 		e.g. "x LIKE 'abc*'" => "x>='abc' AND x<'abd'" etc.

    14 

    15 	Query complexity-reduction by removing redundant predicates, e.g.

    16 		"x>4 AND x>3" => "x>4"

    17 		"x>4 OR x>3" => "x>3"

    18 		"x>-1" for unsigned x => trivially true

    19 

    20 	Better statistical data dn guesswork

    21 

    22 SQL/access plans:

    23 	

    24 	Support "DISTINCT" in the projection stage via a Hash-system or sorting

    25 

    26 

    27