[20181122]了解数据库缓存被那些对象占用.txt

[20181122]了解数据库缓存被那些对象占用.txt --//没事,随手写一个脚本,看看数据库缓存被那些对象占用: column object_name format a30 column owner format a20 column number_MB format 99999999.99 SELECT *   FROM (  SELECT o.owner,                  o.object_name,                  COUNT (*) number_of_blocks,                  COUNT (DISTINCT FILE# || '.' || BLOCK#) distinct_block_count, sum (decode(dirty,'Y',1,0)) dirty_block,                  COUNT (*) - COUNT (DISTINCT FILE# || '.' || BLOCK#)                     diff_number_block,                  ROUND (COUNT (*) * 8 / 1024, 2) number_Mb, ROUND ( (ratio_to_report (SUM (1)) OVER () * 100),2) rr             FROM dba_objects o, v$bh v            WHERE o.data_object_id = v.objd AND o.owner != 'SYS' and v.status <> 'free'         GROUP BY o.owner, o.object_name         ORDER BY COUNT (*) DESC)  WHERE ROWNUM <= 50; --//我直接按照8k数据块计算.取前50个对象. --//顺便提一下,对于cluster table对象无效.不过一般应用很少应用这种类型的表. --//把条件o.owner != 'SYS'改写成o.owner = 'SYS',就明白什么回事. column object_name format a30 column owner format a20 column number_MB format 99999999.99 SELECT *   FROM (  SELECT o.owner,                  o.object_name,                  v.objd,                  COUNT (*) number_of_blocks,                  COUNT (DISTINCT FILE# || '.' || BLOCK#) distinct_block_count,                  sum (decode(dirty,'Y',1,0)) dirty_block,                  COUNT (*) - COUNT (DISTINCT FILE# || '.' || BLOCK#)                     diff_number_block,                  ROUND (COUNT (*) * 8 / 1024, 2) number_Mb, ROUND ( (ratio_to_report (SUM (1)) OVER () * 100),2) rr             FROM dba_objects o, v$bh v            WHERE o.data_object_id = v.objd AND o.owner = 'SYS' and v.status <> 'free'         GROUP BY o.owner, o.object_name,v.objd         ORDER BY COUNT (*) DESC)  WHERE ROWNUM <= 50;