Oracle 11g alert log 新增訊息 opiodr aborting process unknown ospid (1951) as a result of ORA-28 說明

 

一．Alert log 說明

 

在Oracle 11.2.0.2 的資料庫中alert log發現大量如下log：

 

Fatal NI connect error 12537, connectingto:

 (LOCAL=NO)

 

VERSION INFORMATION:

       TNS for HPUX: Version 11.2.0.2.0 - Production

       Oracle Bequeath NT Protocol Adapter for HPUX: Version 11.2.0.2.0 -Production

       TCP/IP NT Protocol Adapter for HPUX: Version 11.2.0.2.0 - Production

 Time: 11-SEP-2011 17:03:56

 Tracing not turned on.

  Tnserror struct:

  nsmain err code: 12537

   

TNS-12537: TNS:connection closed

   ns secondary err code: 12560

   nt main err code: 0

   nt secondary err code: 0

   nt OS err code: 0

opiodr aborting process unknown ospid (29942) as a result of ORA-609

Sun Sep 11 17:05:23 2011

 

 

Tue Sep 13 17:56:37 2011

opiodr aborting processunknown ospid (924) as a result of ORA-28

Tue Sep 13 17:56:38 2011

opiodr aborting process unknownospid (1951) as a result of ORA-28

Tue Sep 13 17:56:40 2011

opiodr aborting process unknown ospid(1983) as a result of ORA-28

Tue Sep 13 17:56:42 2011

opiodr aborting process unknown ospid(11149) as a result of ORA-28

Tue Sep 13 17:56:43 2011

opiodr aborting process unknown ospid(17662) as a result of ORA-28

Tue Sep 13 17:58:06 2011

 

 

opidcl aborting process unknown ospid(9702) as a result of ORA-604

 

二．問題分析：

 

2.1 opiodr aborting process unknown ospid (924) as a resultof ORA-28

 

在Oracle 11.1.0.6 和 11.1.0.7 的版本中，存在一個bug：6377339。 其會在alert log 顯示如下資訊：

 

 ORA-28 : opiodr aborting process unknown ospid (21016_3086862016)

 

       這樣的提示資訊會讓我們產生疑惑. 所以在Oracle 11.2 版本中對這個bug進行修複，修複之後的提示資訊變成了：

opidcl abortingprocess unknown ospid (32110) as a result of ORA-28

 

opiodr aborting process資訊僅僅是一個訊息，它是Oracle 11g裡新添加的一個訊息，用來記錄進程的異常退出。

該訊息一般用來記錄有許可權的使用者kill 的session資訊。

 

 

樣本：

"opiodr aborting process unknown ospid(28342) as a result of ORA-28"

解釋：

"unknown" => means it is not abackground or shadow process that is killed
"ospid (28342)" => this is the OS pid of the process which opiodris aborting
"as a result of" => this precedes the error message which is thereason for opiodr to kill the process
"ORA-28" => this is the reason that opiodr killed the process. Inthis case, it is ora-28.

 

 

參考：

"opiodr abortingprocess" Messages in Alert.Log [ID 1230858.1]

Bug 6377339 - Misleading alertlog message for "opiodr aborting" [ID 6377339.8]

 

 

2.2 opiodr aborting process unknown ospid (29942) as a resultof ORA-609

 

這裡表示進程ospid (29942) 因為ORA-609的錯誤而被kill。導致這個錯誤的原因有2種：

（1） Bug9856585 - "audit network by access does not audit ora-12650"

       該bug 存在Oracle 11.1 到 11.2 的版本中，官網對這個bug的說明如下：

The Bug9856585 is going to be fix bythe Oracle RDBMS Development team, it is currently in Work in Progress Status.

 

（2）在啟動 EM Console後報該錯誤，該問題存在Oracle 10.1到11.2.可能是nscd 進程引起。解決方案是重啟這個進程。

 

用超級使用者（root）使用者執行以下步驟進行重啟：

1. Stop nscd daemon:
# service nscd stop 

2. Verify if any rogue processes is still running:

# ps -ef|grep nscd 

If any process is found, kill the roguenscd process

# kill -9 <nscd_pid> 

3. Start nscd daemon:

# service nscd start 

 

4. As the OS user who has installed theAgent, start the Agent:
cd <AGENT_HOME>/bin
emctl start agent 

 

 

參考：

Audit Network Does Not Work When notSuccessful in Oracle 11gR2 [ID 1241325.1]

Grid Control Agent 10.2 Crashes with CoreDump if 'nscd' daemon is Hung on Linux OS [ID 430200.1]

Lsnrctl Status is Failing and Alert.logShows TNS-12537 and ORA-609 [ID 950135.1]

 

 

2.3 opidcl aborting process unknown ospid (9702) asa result of ORA-604

 

根據2.1 中的說明，這裡的資訊時記錄了ospid(9702) 進程因為ORA-604的錯誤而被中斷了。 該訊息僅用來記錄進程異常退出的資訊，是Oracle 11g中新增加的訊息。

 

 

 

 

三．小結：

       在Oracle 11g中，新增了一個新的訊息機制，會在alert log中記錄異常退出的進程。

 

其訊息格式樣本如下：

"opiodr aborting process unknown ospid(28342) as a result of ORA-28"

解釋：

"unknown" => means it is not abackground or shadow process that is killed
"ospid (28342)" => this is the OS pid of the process which opiodris aborting
"as a result of" => this precedes the error message which is thereason for opiodr to kill the process
"ORA-28" => this is the reason that opiodr killed the process. Inthis case, it is ora-28.

 

這個是Oracle 11.2 以後的格式，在11.2 之前的格式：

ORA-28 : opiodr aborting process unknownospid (21016_3086862016)

 

這種格式因為會讓人迷惑，因此該格式被稱為Bug 6377339，並在11.2中修改了這個格式。

 

Alert log中的"opiodr aborting processunknown ospid (28342) as a result of ORA-28"訊息僅僅是一個記錄資訊。最後的ORA-XX的資訊代表導致進程異常退出的原因。

 

 

 

 

 

 

 

 

-------------------------------------------------------------------------------------------------------

著作權，文章允許轉載，但必須以連結方式註明源地址，否則追究法律責任!

Skype: tianlesoftware

QQ:              tianlesoftware@gmail.com

Email:   tianlesoftware@gmail.com

Blog:     http://www.tianlesoftware.com

Weibo: http://weibo.com/tianlesoftware

Twitter: http://twitter.com/tianlesoftware

Facebook: http://www.facebook.com/tianlesoftware

Linkedin: http://cn.linkedin.com/in/tianlesoftware

 

 

-------加群需要在備忘說明Oracle資料表空間和資料檔案的關係，否則拒絕申請----

DBA1 群：62697716(滿);   DBA2 群：62697977(滿)  DBA3 群：62697850(滿)  

DBA 超級群：63306533(滿);  DBA4 群：83829929   DBA5群： 142216823

DBA6 群：158654907    DBA7 群：172855474   DBA總群：104207940