[ORALCE]SQL 优化案例之 索引的聚合因子clustering factor

Posted 2020-12-15 A running snail,little step ev

索引查询要尽可能的避免回表,如不可避免,要关注聚合因子是否过大,聚合因子过大,回表代价高,产生的bufer 多

第一步，构造有序列ｘ在表＂colocated ＂，无序列ｘ在表＂disorganized＂

--构造有序列ｘ在表＂colocated ＂
drop table colocated  purge;
create table colocated (x int,y varchar2(80));
begin
    for i in 1 .. 100000
    loop
        insert into colocated(x,y)
        values (i,rpad(dbms_random.random,75,\'*\'));
    end loop;
end;
/

alter table colocated add constraint colocated_pk primary key (x);
begin
    dbms_stats.gather_table_stats( \'SYS\', \'colocated\', cascade => true);
end;
/

--disorganized 列x完全无序
drop table disorganized purge;
create table disorganized as select x,y from colocated order by y;
alter table disorganized add constraint disorganized_pk primary key(x);
begin
    dbms_stats.gather_table_stats(\'SYS\',\'disorganized\',cascade=>true);
end;
/

第二步 比较

 

 

第三步  查看clustering_factor

col index_name for A15
select
    a.table_name,
    a.index_name,
    a.blevel,
    a.leaf_blocks,
    b.num_rows,
    b.blocks,
    a.clustering_factor,
    trunc(a.clustering_factor / b.num_rows,2) cluster_rate
from dba_indexes a, dba_tables b
where index_name in(\'COLOCATED_PK\',\'DISORGANIZED_PK\') and a.table_name=b.table_name;

TABLE_NAME          INDEX_NAME      BLEVEL LEAF_BLOCKS   NUM_ROWS  BLOCKS   CLUSTERING_FACTOR    CLUSTER_RATE
----------------- --------------- ------ -----------   --------- -------    ------------------   ------------
COLOCATED          COLOCATED_PK        1     208           100000     1191           1190                 .01
DISORGANIZED       DISORGANIZED_PK     1     208           100000     1191          99913                 .99