数据库中日志基本不切换，db基本没有负载，查看警告日志中发现在出现了ora-00600后，redo的切换就不正常了，这里也提醒下数据库没有问题不代表真的没问题，对于系统有些异常情况需要及时去查看后台日志来诊断具体的信息。系统主机也没有什么负载。

Thu May 3 12:29:28 2012

ORA-00060: Deadlock detected. More info in file /db/oracle10g/admin/benguo/udump/benguo_ora_21458.trc.

Thu May 3 12:41:22 2012

ORA-00060: Deadlock detected. More info in file /db/oracle10g/admin/benguo/udump/benguo_ora_21430.trc.

Thu May 3 12:56:11 2012

ORA-00060: Deadlock detected. More info in file /db/oracle10g/admin/benguo/udump/benguo_ora_21568.trc.

Thu May 3 13:29:26 2012

ORA-00060: Deadlock detected. More info in file /db/oracle10g/admin/benguo/udump/benguo_ora_21469.trc.

Thu May 3 13:36:04 2012

Errors in file /db/oracle10g/admin/benguo/udump/benguo_ora_21650.trc:

ORA-00600: internal error code, arguments: [12333], [19], [3], [15], [], [], [], []

Thu May 3 14:01:20 2012

Thread 1 advanced to log sequence 4402

Current log# 4 seq# 4402 mem# 0: /db/oracle10g/oradata/benguo/redo04.log

Thu May 3 15:28:13 2012

Errors in file /db/oracle10g/admin/benguo/udump/benguo_ora_21646.trc:

ORA-00600: internal error code, arguments: [12333], [19], [3], [15], [], [], [], []

Thu May 3 18:01:25 2012

Thread 1 advanced to log sequence 4403

Current log# 5 seq# 4403 mem# 0: /db/oracle10g/oradata/benguo/redo05.log

Trace文件

*** SERVICE NAME:(SYS$USERS) 2012-05-04 14:06:53.162

*** SESSION ID:(412.37869) 2012-05-04 14:06:53.162

*** 2012-05-04 14:06:53.162

ksedmp: internal or fatal error

ORA-00600: internal error code, arguments: [12333], [19], [3], [94], [], [], [], []

Current SQL statement for this session:

SELECT COUNT(f_article_id) AS f_article_id FROM textauto_folderarticle WHERE f_folder_id = :1

----- Call Stack Trace -----

calling call entry argument values in hex

location type point (? means dubious value)

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

ksedst()+31 call ksedst1() 000000000 ? 000000001 ?

000000000 ? 000000000 ?

000000000 ? 000000001 ?

ksedmp()+610 call ksedst() 000000000 ? 000000001 ?

000000000 ? 000000000 ?

000000000 ? 000000001 ?

ksfdmp()+21 call ksedmp() 000000003 ? 000000001 ?

000000000 ? 000000000 ?

000000000 ? 000000001 ?

kgeriv()+176 call ksfdmp() 000000003 ? 000000001 ?

000000000 ? 000000000 ?

000000000 ? 000000001 ?

kgesiv()+119 call kgeriv() 005EBDCA0 ? 2AEA455E2160 ?

000000001 ?

根据该trace文件是这个sql语句的绑定变量后导致的错误，由于网络上没有过多的解释，采取的重启java进程，后续这个错误就没有发生了，日志也切换正常了！

http://www.itpub.net/thread-1422637-1-1.html

http://www.itpub.net/thread-249828-2-1.html

的ITPUB中帖子中也出现了同样的问题，关于其中的缘由网友给出的解答貌似是个bug导致，不过该内部错误是否导致了系统日志切换问题应该不太可能，因为在同样的正常的日志切换的数据库中依然发现此类错误，更可能相信是一个不影响db的小bug，而且出现问题的数据库正是10.2.0.1.0版本，在oracle 11g下已经解决

db: 10.2.0.0--->10.2.0.3.0
os: all

Metalink DocID: Note:464321.1

Stack trace is similar to:

ksedst ksedmp ksfdmp kgeriv kgesiv ksesic3
opitsk opiino opiodr opidrv sou2o opimai_real
main $START$


Cause:
This is Bug 5926074
Abstract: ORA-600 [12333] UPDATE CLOB

If you update on two clob columns and you get ora-600[12333], then you hit this
bug.

solution:

a. fixed in Oracle server version 11.0

b. get the patch: 5926074 for you specify os

[@more@]