监控
1.监控事例的等待：
Selectevent,sum(decode(wait_time,0,0,1))prev,sum(decode(wait_time,0,1,0))curr,count(*)fromv$session_waitgroupbyenentorderby4;
2.回滚段的征用情况：
Selectname,waits,gets,waits/getsrationfromv$rollstata,v$rollnamebwherea.usn=b.usn;
3.监控表空间I/O比例：
Selectdf.tablespace_namename,df.file_name“file”,f.phyrdspyr,f.phyblkrdpdr,f.phywrtspyw,
f.phyblkwrtpbwfromv$filestatf,dba_data_filesdfwheref.file#=df.file_id
4.监控文件系统的I/O比例：
Selectsubstr(a.file#,1,2)”#”,substr(a.name,1,30)”name”,a.status,a.bytes,
b.phyrds,b.phywrtsfromv$datafilea,v$filestatbwherea.file#b.file#
5.在某个用户下找所有的索引：
Selectuser_indexes.table_name,user_indees.index_name,uniqueness,column_name
Fromuser_ind_columns,user_indexesc
Whereuser_ind_columns.inde_name=user_indexes.index_name
anduser_ind_columns.table_name=user_indexes.table_name
orderbyuser_indexes.table_type,user_indexes.table_name,
user_indexes.index_name,column_position;
6.监控SGA的命中率
Selecta.value+b.value“logical_reads”,c.value“phys_reads”,
Round(100*((a.value+b.value)-c.value)/(a.value+b.value)):BUFFERHITRATIO”
Fromv$sysstata.v$sysstatb,v$sysstatc
Wherea.statistic#=38andb.statistic#=39andc.statistic#=40;
7.监控SGA中字典缓冲区的命中率
Selectparameter,gets,Getmisses,getmisses/(gets+getmisses)*100“missratio”,
(1-(sum(getmisses)/(sum(gets)+sum(getmisses))))*100”Hitratio”
Fromv$rowcachewheregets+getmisses<>0
Groupbyparameter,gets,getmisses;
8.监控SGA中共享缓冲区的命中率，应该小于1%
Selectsum(pins)”TotalPins”,sum(reloads)“TotalReloads”,
Sum(reloads)/sum(pins)*100libcachefromv$librarycache;
Selectsum(pinhits-reloads)/sum(pins)”hitradio”,sum*reloads)/sum(pins)”reloadpercent”
Fromv$librarycache;
9.显示所有数据库对象的类别和大小
Selectcount(name)num_instances,type,sum(source_size)source_size,
Sum(parsed_size)parsed_size,sum(code_dize)code_dize,sum(error_size)error_size,
Sum(source_size)+sum(parsed_size)+sum(code_size)+sum(error_size)size_required
Fromdba_object_sizegroupbytypeororderby2;
10.监控SGA中重做日志缓冲区的命中率，应小于1%
Selectname,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
FROMv$latchWHEREnameIN(‘redoallocation’,’redocopy’);
11.监控内存和硬盘的排列比率，最好使他小于增加说sort_area_sze
Selectname,valuefromv$sysstatwherenamein(‘sorts(memory)’,’sorts(disk)’);
12.监控当前数据库谁在运行什么SQL语句
Selectosuser,username,sql_textfromv$sessiona,v$sqltextb
Wherea.sql_address=b.addressorderbyaddress,piece;
13.监控字典缓冲区
Select(sum(pins-reloads))/sum(pins)”libcache”fromv$librarycache;
Select(sum(gets-getmisses-usage-fixed))/sum(gets)”rowcache”fromv$rowcache;
Selectsum(pins)”executions”,sum(reloads)”cachemisseswhileexecuting”fromv$librarycache;
后者除以前者，此比率小于1%接近0%为好。
Selectsum(gets)”dictionarygets”,sum(getmisses)”dictionarycachegetmisses”
Fromv$rowcache
14.找ORACE字符集
Select*fromsys.props$wherename=’NLS_CHARACTERSET’;
15.监控MTS
Selectbusy/(busy+idle)”sharedserversbusy”fromv$dispatcher;
此值大于0.5时，参数需加大
Selectsum(wait)/sum(totalq)”dispatcherwaits”fromv$queuewheretpye=’dispatcher’;
Selectcount(*)fromv$dispatcher;
Selectservers_highwaterfromv$mts;
Severs_highwater接近mts_max_servers时，参数需加大
16.碎片程度
Selecttablespace_name,count(tablespace_name)fromdba_spacegroupbytablespace_name
Havingcount(tablespace_name)>10;
Altertablespacenamecoalesce;
Altertablenamedeallocateunused;
Createorreplaceviewts_blocks_vas
Selecttablespace_name,block_id,bytes,blocks,’freespace’segment_namefromdba_free_spaceunionall
Selecttablespace_name,block_id,bytes,blocks,segment_namefromdba_extents;
Select*fromts_blocks_v;
Selecttablespace_name,sum(bytes),ma(bytes),count(block_id)fromdba_free_spacegroupbytablespace_name;
查看碎片程度高的表
SELECTsegment_nametable_name,COUNT(*)extents
FROMdba_segementsWHEREownerNOTIN(‘SYS’,’SYSTEM’)GROUPBYgegment_name
HAVINGCOUNT(*)=(SELECTMAX(COUNT(*)FROMdba_segmentsGROUPBYsegment_name);
17.表、索引的存储情况检查
Selectsement_name,sum(bytes),count(*)_quanfromdba_extentswhere
Tablespace_name=’&tablespace_name’andsegment_type=’TABLE’groupbytablesapce_name,segment_name;
Selectsegment_name,count(*)fromdba_extentswheresegment_type=’INDEX’andowner=’&owner’groupbysegment_name;j