ORA-609：疑似MEMORY_TARGET設定過大導致的宕機

ORA-609：疑似MEMORY_TARGET設定過大導致的宕機

現象：
客戶的資料庫最近發生好幾次忽然宕機事件，查看alert日誌：
Thu Oct 10 10:47:50 2013
ORA-609 : opiodr aborting process unknown ospid (834_47932399713424)
Thu Oct 10 10:47:50 2013
ORA-609 : opiodr aborting process unknown ospid (740_47613956914320)
Thu Oct 10 10:47:53 2013 
ORA-609 : opiodr aborting process unknown ospid (755_47556191542416)
Thu Oct 10 10:47:53 2013
ORA-609 : opiodr aborting process unknown ospid (761_47545590688912)
Thu Oct 10 10:47:54 2013 
ORA-609 : opiodr aborting process unknown ospid (859_47653507185808)
Thu Oct 10 10:47:58 2013
ORA-609 : opiodr aborting process unknown ospid (904_47121713061008)
Thu Oct 10 10:48:08 2013
ORA-609 : opiodr aborting process unknown ospid (918_47924605709456)
Thu Oct 10 10:48:16 2013
ORA-609 : opiodr aborting process unknown ospid (965_46996288695440)
Thu Oct 10 10:48:16 2013 
ORA-609 : opiodr aborting process unknown ospid (984_47202090368144)
Thu Oct 10 10:49:06 2013
ORA-609 : opiodr aborting process unknown ospid (1103_47027913133200)
Thu Oct 10 10:49:08 2013
Thu Oct 10 10:49:08 2013
ORA-609 : opiodr aborting process unknown ospid (1064_47991325715600)
ORA-609 : opiodr aborting process unknown ospid (1150_47885260754064)
Thu Oct 10 10:50:02 2013
Process W000 died, see its trace file
Thu Oct 10 10:50:17 2013
Process W001 died, see its trace file
Thu Oct 10 10:54:19 2013
Process W000 died, see its trace file
Process W000 died, see its trace file
Process W000 died, see its trace file
Thu Oct 10 10:54:31 2013
Process W000 died, see its trace file
Process W000 died, see its trace file
Process W000 died, see its trace file
Thu Oct 10 10:54:43 2013
Process W000 died, see its trace file
Process W000 died, see its trace file
Process W000 died, see its trace file
Thu Oct 10 10:54:55 2013
Process W000 died, see its trace file
Process W000 died, see its trace file
Process W000 died, see its trace file
Thu Oct 10 10:55:07 2013
Process W000 died, see its trace file
Process W000 died, see its trace file
Process W000 died, see its trace file
Thu Oct 10 10:55:19 2013
Process W000 died, see its trace file
Process W000 died, see its trace file
Process W000 died, see its trace file
Thu Oct 10 10:55:31 2013
Process W000 died, see its trace file
Process W000 died, see its trace file
Process W000 died, see its trace file
Thu Oct 10 11:02:06 2013
Thread 2 advanced to log sequence 18269
 Current log# 15 seq# 18269 mem# 0: +UNIDG/uni/onlinelog/group_15.295.796513629
Thu Oct 10 11:02:08 2013
SUCCESS: diskgroup UNIDG was mounted
Thu Oct 10 11:02:18 2013
SUCCESS: diskgroup UNIDG was dismounted
Thu Oct 10 11:03:40 2013
Process m001 died, see its trace file
Thu Oct 10 11:04:23 2013
Process m001 died, see its trace file

分析：
查看啟動時的參數設定，發現客戶的系統啟用了自動記憶體管理：
memory_target = 21G <-------------Automatic Memory Management feature is used.
free -g查看記憶體為23G
懷疑主機資源耗盡導致系統宕機，建議客戶memory_target設定為16-18G。
調小memort_target之後，之前的自動宕機現象消失。

附：分析了如下日誌。
$CRS_HOME/log/<nodename>/*.log
$CRS_HOME/log/<nodename>/crsd/*.log
$CRS_HOME/log/<nodename>/cssd/*.log
/etc/oracle/oprocd/*.log.* or /var/opt/oracle/oprocd/*.log.*