1、普通的压缩表
SQL> create table uncompressed pctfree 0 as select * from dba_objects order by owner, object_type, object_name;
Table created.
SQL> create table compressed COMPRESS as select * from uncompressed order by owner, object_type, object_name;
Table created.
SQL> analyze table compressed compute statistics for table;
Table analyzed.
SQL> analyze table uncompressed compute statistics for table;
Table analyzed.
SQL> select cblks comp_blks, uncblks uncomp_blks,round(cblks/uncblks*100,2) pct
from (select max(decode(table_name,'COMPRESSED',blocks,null)) cblks,
max(decode(table_name,'UNCOMPRESSED',blocks,null)) uncblks
from user_tables where table_name in ( 'COMPRESSED', 'UNCOMPRESSED' )
5 );
COMP_BLKS UNCOMP_BLKS PCT
---------- ----------- ----------
328 618 53.07 看来效果一般
2 随机压缩表
SQL> drop table compressed;
Table dropped.
SQL> create table compressed COMPRESS as select * from uncompressed order by dbms_random.random;
Table created.
SQL> analyze table compressed compute statistics for table;
Table analyzed.
SQL> select cblks comp_blks, uncblks uncomp_blks,round(cblks/uncblks*100,2) pct
from (select max(decode(table_name,'COMPRESSED',blocks,null)) cblks,
max(decode(table_name,'UNCOMPRESSED',blocks,null)) uncblks
from user_tables where table_name in ( 'COMPRESSED', 'UNCOMPRESSED' )
5 );
COMP_BLKS UNCOMP_BLKS PCT
---------- ----------- ----------
444 618 71.84 效果比较差
3、基于时间戳的压缩表
SQL> drop table compressed;
Table dropped.
SQL> create table compressed COMPRESS as select * from uncompressed order by timestamp;
Table created.
SQL> analyze table compressed compute statistics for table;
Table analyzed.
SQL> select cblks comp_blks, uncblks uncomp_blks,round(cblks/uncblks*100,2) pct
from (select max(decode(table_name,'COMPRESSED',blocks,null)) cblks,
max(decode(table_name,'UNCOMPRESSED',blocks,null)) uncblks
from user_tables where table_name in ( 'COMPRESSED', 'UNCOMPRESSED' )
5 );
COMP_BLKS UNCOMP_BLKS PCT
---------- ----------- ----------
218 618 35.28 效果很不错
来自 “ ITPUB博客 ” ,链接:http://blog.itpub.net/21158541/viewspace-668715/,如需转载,请注明出处,否则将追究法律责任。
转载于:http://blog.itpub.net/21158541/viewspace-668715/