原来ora-600并不是最可怕的

原来ora-600并不是最可怕的，可怕的是造成它的根源[@more@]

一网友数据库alert报如下错误信息：

Timed out trying to start process J000.
Wed Jul 30 16:16:05 2008
Timed out trying to start process J000.
Wed Jul 30 16:19:14 2008
Errors in file /home/oracle/admin/rac/bdump/rac4_pmon_22083.trc:
ORA-00600: internal error code, arguments: [OSDEP_INTERNAL], [], [], [], [], [], [], []
ORA-27302: failure occurred at: skgslfr
Wed Jul 30 16:19:17 2008
Errors in file /home/oracle/admin/rac/bdump/rac4_pmon_22083.trc:
ORA-00600: internal error code, arguments: [513], [], [], [], [], [], [], []
ORA-00600: internal error code, arguments: [OSDEP_INTERNAL], [], [], [], [], [], [], []
ORA-27302: failure occurred at: skgslfr
Wed Jul 30 16:19:17 2008
Errors in file /home/oracle/admin/rac/bdump/rac4_pmon_22083.trc:
ORA-00600: internal error code, arguments: [513], [], [], [], [], [], [], []
ORA-00600: internal error code, arguments: [OSDEP_INTERNAL], [], [], [], [], [], [], []
ORA-27302: failure occurred at: skgslfr
Wed Jul 30 16:19:17 2008
PMON: terminating instance due to error 600

从alert看，是J000启动失败，遭遇600，pmon进程终止了实例。

读完trace文件，发现超过100个的死进程，在trace文件末尾发现写sq enqueue的信息。

询问，原来开始实例hang住了，登录不进去，dba也进不去。于是手动kill掉了许多进程，，结果实例down了，重启ok了，该问题出现不止一次。hang时alert中记录回滚段疯狂增长，session也增加很多。环境是linux+9204，应该是4节点rac环境。

于是乎，觉得600是因为进程被kill掉导致的，没处理的必要了，转到实例为什么hang住了，已经hang时表现的一些异常现象。

问题就先打住吧，要说的是原来ora-600并不是最可怕的，可怕的是造成它的根源。

从改问题的第一经手人，从改问题的第二经手人角度，都有些要思考的问题啊，:-)

最终原因不能肯定，已知的一个因素是audses$ 的cachesize，相关资料也记录下吧：

PURPOSE
-------

Prevent hangs in RAC due to high login rate and low cache setting of AUDSES$ sequence.
Default of 20 is not adequate in many cases.

PROBLEM:
--------

The default setting for the SYS.AUDSES$ sequence is 20, this is too low for a RAC system
where logins can occur concurrently from multiple nodes. During high login rate
such low value can cause slowness and even hangs. Some of the symptoms are:
- Checkpoint not completing on all RAC nodes
- Waits expire on row cache enqueue lock dc_sequences
- RAC hangs due to QMON deadlocking

During those hangs session login is not possible (or *extremely* slow) due
to extremely high contention on the above sequence.

SOLUTION:
---------

Manually increase that sequence cache on each affected database:

alter sequence sys.audses$ cache 10000;

This is fixed in 10.2.0.3 patchset. (Affected releases 9i to 10.2.0.2)

来自 “ ITPUB博客 ” ，链接：http://blog.itpub.net/7591490/viewspace-1008198/，如需转载，请注明出处，否则将追究法律责任。

转载于:http://blog.itpub.net/7591490/viewspace-1008198/