Oracle监控的关键指标(转)

1、监控事例的等待

select event,sum(decode(wait_Time,0,0,1)) "Prev",     
sum(decode(wait_Time,0,1,0)) "Curr",count(*) "Tot"     
from v$session_Wait     
group by event order by 4;    
2、回滚段的争用情况

select name, waits, gets, waits/gets "Ratio"     
from v$rollstat a, v$rollname b     
where a.usn = b.usn;    
3、监控表空间的 I/O 比例

select df.tablespace_name name,df.file_name "file",f.phyrds pyr,    
f.phyblkrd pbr,f.phywrts pyw, f.phyblkwrt pbw    
from v$filestat f, dba_data_files df    
where f.file# = df.file_id    
order by df.tablespace_name;    
4、监控文件系统的 I/O 比例

select substr(a.file#,1,2) "#", substr(a.name,1,30) "Name",     
a.status, a.bytes, b.phyrds, b.phywrts     
from v$datafile a, v$filestat b     
where a.file# = b.file#;     
5、在某个用户下找所有的索引

select user_indexes.table_name, user_indexes.index_name,uniqueness, column_name    
from user_ind_columns, user_indexes    
where user_ind_columns.index_name = user_indexes.index_name    
and user_ind_columns.table_name = user_indexes.table_name     
order by user_indexes.table_type, user_indexes.table_name,    
user_indexes.index_name, column_position;    
6、监控 SGA 的命中率

select a.value + b.value "logical_reads", c.value "phys_reads",    
round(100 * ((a.value+b.value)-c.value) / (a.value+b.value)) "BUFFER HIT RATIO"     
from v$sysstat a, v$sysstat b, v$sysstat c    
where a.statistic# = 38 and b.statistic# = 39     
and c.statistic# = 40;     
7、监控 SGA 中字典缓冲区的命中率

select parameter, gets,Getmisses , getmisses/(gets+getmisses)*100 "miss ratio",    
(1-(sum(getmisses)/ (sum(gets)+sum(getmisses))))*100 "Hit ratio"    
from v$rowcache     
where gets+getmisses <>0    
group by parameter, gets, getmisses;     
8、监控 SGA 中共享缓存区的命中率，应该小于1%

select sum(pins) "Total Pins", sum(reloads) "Total Reloads",    
sum(reloads)/sum(pins) *100 libcache    
from v$librarycache;    

select sum(pinhits-reloads)/sum(pins) "hit radio",sum(reloads)/sum(pins) "reload percent"     
from v$librarycache;    
9、显示所有数据库对象的类别和大小

select count(name) num_instances ,type ,sum(source_size) source_size ,    
sum(parsed_size) parsed_size ,sum(code_size) code_size ,sum(error_size) error_size,    
sum(source_size) +sum(parsed_size) +sum(code_size) +sum(error_size) size_required     
from dba_object_size     
group by type order by 2;    
10、监控 SGA 中重做日志缓存区的命中率，应该小于1%

SELECT name, gets, misses, immediate_gets, immediate_misses,    
Decode(gets,0,0,misses/gets*100) ratio1,    
Decode(immediate_gets+immediate_misses,0,0,    
immediate_misses/(immediate_gets+immediate_misses)*100) ratio2    
FROM v$latch WHERE name IN ('redo allocation', 'redo copy');     
11、监控内存和硬盘的排序比率，最好使它小于 .10，增加 sort_area_size

SELECT name, value FROM v$sysstat WHERE name IN ('sorts (memory)', 'sorts (disk)');     
12、监控当前数据库谁在运行什么SQL语句

SELECT osuser, username, sql_text from v$session a, v$sqltext b    
where a.sql_address =b.address order by address, piece;    
13、监控字典缓冲区

SELECT (SUM(PINS - RELOADS)) / SUM(PINS) "LIB CACHE" FROM V$LIBRARYCACHE;    
SELECT (SUM(GETS - GETMISSES - USAGE - FIXED)) / SUM(GETS) "ROW CACHE" FROM V$ROWCACHE;    
SELECT SUM(PINS) "EXECUTIONS", SUM(RELOADS) "CACHE MISSES WHILE EXECUTING" FROM V$LIBRARYCACHE;     

后者除以前者，此比率小于1%，接近0%为好。    

SELECT SUM(GETS) "DICTIONARY GETS",SUM(GETMISSES) "DICTIONARY CACHE GET MISSES"    
FROM V$ROWCACHE    
14、找ORACLE字符集

select * from sys.props$ where name='NLS_CHARACTERSET';     
15、监控 MTS

select busy/(busy+idle) "shared servers busy" from v$dispatcher;    

此值大于0.5时，参数需加大    

select sum(wait)/sum(totalq) "dispatcher waits" from v$queue where type='dispatcher';    
select count(*) from v$dispatcher;    
select servers_highwater from v$mts;    

servers_highwater接近mts_max_servers时，参数需加大    
16、碎片程度

select tablespace_name,count(tablespace_name) from dba_free_space group by tablespace_name     
having count(tablespace_name)>10;    

alter tablespace name coalesce;    
alter table name deallocate unused;    

create or replace view ts_blocks_v as    
select tablespace_name,block_id,bytes,blocks,'free space' segment_name from dba_free_space    
union all    
select tablespace_name,block_id,bytes,blocks,segment_name from dba_extents;    

select * from ts_blocks_v;    

select tablespace_name,sum(bytes),max(bytes),count(block_id) from dba_free_space     
group by tablespace_name;    

查看碎片程度高的表    

SELECT segment_name table_name , COUNT(*) extents    
FROM dba_segments WHERE owner NOT IN ('SYS', 'SYSTEM') GROUP BY segment_name    
HAVING COUNT(*) = (SELECT MAX( COUNT(*) ) FROM dba_segments GROUP BY segment_name);    
17、表、索引的存储情况检查

select segment_name,sum(bytes),count(*) ext_quan from dba_extents where     
tablespace_name='&tablespace_name' and segment_type='TABLE' group by tablespace_name,segment_name;    

select segment_name,count(*) from dba_extents where segment_type='INDEX' and owner='&owner'    
group by segment_name;    
18、找使用CPU多的用户session

select a.sid,spid,status,substr(a.program,1,10) prog,a.terminal,osuser,value/60/100 value    
from v$session a,v$process b,v$sesstat c    
where c.statistic#=12 and c.sid=a.sid and a.paddr=b.addr order by value desc;