ORA-27061: skgfospo: waiting for async I/Os failed

author：skate
time：2010/02/26

---

 

ORA-27061: skgfospo: waiting for async I/Os failed

 

RMAN-00571: ===========================================================
RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS ===============
RMAN-00571: ===========================================================
RMAN-00601: fatal error in recovery manager
RMAN-03004: fatal error during execution of command
RMAN-10039: error encountered while polling for RPC completion on channel default
RMAN-10006: error running SQL statement: select act from x$ksusex where sid=:1 and serial=:2
RMAN-10002: ORACLE error: ORA-03113: end-of-file on communication channel
ORA-03114: not connected to ORACLE
RMAN-03002: failure of recover command at 03/02/2010 07:15:30
RMAN-11003: failure during parse/execution of SQL statement: alter database recover logfile '/u02/backup/1_31141.dbf'
ORA-03113: end-of-file on communication channel

alert.log文件内容：

 

Media Recovery Log
ORA-279 signalled during: alter database recover if needed
 start until canc...
Tue Mar  2 07:15:09 2010
alter database recover logfile '/u02/backup/1_31140.dbf'
Tue Mar  2 07:15:09 2010
Media Recovery Log /u02/backup/1_31140.dbf
ORA-279 signalled during: alter database recover logfile '/u02/backup/1_3114...
Tue Mar  2 07:15:16 2010
alter database recover logfile '/u02/backup/1_31141.dbf'
Media Recovery Log /u02/backup/1_31141.dbf
Tue Mar  2 07:15:26 2010
Errors in file /home/oracle/admin/orcl/bdump/orcl_dbw0_369106.trc:
ORA-27061: skgfospo: waiting for async I/Os failed
IBM AIX RISC System/6000 Error: 22: Invalid argument
Tue Mar  2 07:15:26 2010
DBW0: terminating instance due to error 27061
Tue Mar  2 07:15:28 2010
Errors in file /home/oracle/admin/orcl/bdump/orcl_lgwr_352716.trc:
ORA-27061: skgfospo: waiting for async I/Os failed
Tue Mar  2 07:15:29 2010
Errors in file /home/oracle/admin/orcl/bdump/orcl_ckpt_438366.trc:
ORA-27061: skgfospo: waiting for async I/Os failed
Tue Mar  2 07:15:29 2010
Errors in file /home/oracle/admin/orcl/bdump/orcl_pmon_319782.trc:
ORA-27061: skgfospo: waiting for async I/Os failed
Tue Mar  2 07:15:39 2010
Instance terminated by DBW0, pid = 369106

# more /home/oracle/admin/orcl/bdump/orcl_ckpt_438366.trc
/home/oracle/admin/orcl/bdump/orcl_ckpt_438366.trc
Oracle9i Enterprise Edition Release 9.2.0.1.0 - 64bit Production
With the Partitioning, OLAP and Oracle Data Mining options
JServer Release 9.2.0.1.0 - Production
ORACLE_HOME = /home/oracle/product/9.2.0
System name:    AIX
Node name:      db1
Release:        3
Version:        5
Machine:        00C64E4B4C00
Instance name: orcl
Redo thread mounted by this instance: 1
Oracle process number: 5
Unix process pid: 438366, image: oracle@db1 (CKPT)

*** 2010-03-02 07:15:29.155
*** SESSION ID:(4.1) 2010-03-02 07:15:29.147
error 27061 detected in background process
ORA-27061: skgfospo: waiting for async I/Os failed

 

oracle版本：Oracle9i Enterprise Edition Release 9.2.0.1.0
os版本：AIX5.3-TL3

 

这个由于异步IO引起的问题，在AIX 5.3 TL3下，使用文件系统，打开Oracle的异步IO选项的时候，经常会出现。
当异步IO的请求（队列）超出MAXREQS的时候，由于AIX上AIOWAIT的问题，会导致返回值是错误的，Oracle发现
该错误后，就会出现问题。如果正好是某些重要的后台进程碰到这样的错误，会导致数据库宕掉。

 

解决方法（三选一，其中第二和第三可以一起使用）：

1、关闭数据库的异步IO（最后没办法的时候采取的措施）

 

2、修改异步IO参数，加大MAXREQS到8192，甚至16384（很多人说这个调整无效）

chdev -l aio0 -a maxreqs=16384 --或者8192

要不启动系统，动态修改请使用下面命令（临时设置，重启不起作用）

aioo -o maxreqs=16384   --或者8192  （aioo是AIX5.3-TL5版本以上才支持的）

 

3、安装IBM的IZ03260补丁包（这个是最根本的解决方法，修复bug）

现象：由于AIO的异常，在系统会看见cpu有异常等待，及时系统空闲，cpu也有等待，这个时候的临时解决方法是重启下服务器，让其aio正常工作，找个时间重启下机器

参考文献：

http://www.oraclefans.cn/forum/showtopic.jsp?rootid=5736&CPages=1

http://q.sohu.com/forum/5/topic/1556151

 

 

-----end---