错误信息为：

Errors in file /u1/oracle/admin/repdb01/udump/repdb01_ora_10600.trc:
ORA-07445: exception encountered: core dump [0000000100A2E0F4] [SIGSEGV] [Address not mapped to object] [0x000000008] [] []

对应的trace文件信息为：

ksedmp: internal or fatal error
ORA-07445: exception encountered: core dump [0000000100A2E0F4] [SIGSEGV] [Address not mapped to object] [0x000000008] [] []
Current SQL statement for this session:
INSERT INTO SYS.JOB$ ( JOB,LOWNER,POWNER,COWNER,NEXT_DATE,INTERVAL#,FLAG,WHAT,NLSENV,ENV )  VALUES ( :b1,:b2,:b3,:b4,NVL(:b5,TO_DATE
('4000-01-01','YYYY-MM-DD')),NVL(:b6,'null'),:b7,:b8,:b9,:b10  )
----- PL/SQL Call Stack -----
  object      line  object
  handle    number  name
69817eee0       210  package body SYS.DBMS_IJOB
6aba32d10       150  package body SYS.DBMS_JOB
6ae481538         2  anonymous block
----- Call Stack Trace -----
calling              call     entry                argument values in hex     
location             type     point                (? means dubious value)    
-------------------- -------- -------------------- ----------------------------
ksedmp()+328         CALL     ksedst()+0           FFFFFFFF7FFF5630 ?
                                                   000000000 ? 000000000 ?
                                                   00000003E ?
                                                   FFFFFFFF7FFF5EC8 ?
                                                   1031D56C8 ?
ssexhd()+604         CALL     ksedmp()+0           000000000 ? 000103400 ?
                                                   0001035D9 ? 000102C00 ?
                                                   1035D9000 ? 1035D9C28 ?
sigacthandler()+44   PTR_CALL 0000000000000000     1035E1000 ?
                                                   FFFFFFFF7FFF6E60 ?
                                                   000000000 ? 000000001 ?
                                                   1035DEDD8 ? 00000000B ?
kxscbb()+84          PTR_CALL 0000000000000000     00000000B ?
                                                   FFFFFFFF7FFF6E60 ?
                                                   FFFFFFFF7FFF6B80 ?
                                                   00000000B ? 000000076 ?
                                                   000000000 ?
kxspoac()+620        CALL     kxscbb()+0           FFFFFFFF7CE7CB80 ?
                                                   000000000 ?
                                                   FFFFFFFF7CE7B8C8 ?
                                                   000000018 ? 000000FA0 ?
                                                   000000000 ?
opipls()+2824        CALL     kxspoac()+0          FFFFFFFF7CEAD340 ?
                                                   FFFFFFFF7CE7B8C8 ?
                                                   1036B53D8 ?
                                                   FFFFFFFF7CE7CB80 ?
                                                   000000010 ? 69E4A2CC8 ?
opiodr()+1692        PTR_CALL 0000000000000000     00000000A ? 000000001 ?
                                                   FFFFFFFF7CF7C2E8 ?
                                                   00000000A ? 000000017 ?
                                                   1035D9FC8 ?

错误发生在Oracle将JOB信息插入到JOB$表中，这个语句应该是用户执行DBMS_JOB.SUBMIT过程触发的。

检查随后的日志信息，发现是同事通过PLDEVELOPER工具在运行一个JOB。由于同事已经下班，因此只是电话简单沟通了一下，并确认JOB已经运行成功。看来这个bug的影响不是很大。

随后查询了一下metalink，看了很多的kxscbb参数相关的错误描述，感觉问题可能与Doc ID:  268505.1或Doc ID:  3315960.8文档中描述的比较接近。但是由于不清楚当时出错时的具体信息，因此很难下结论。

等到今天碰到同事详细的问了一下当时的情况，就基本上断定了正是上面两篇文章中描述的情况。由于要执行一个一次性的JOB，因此同事选择使用匿名块的方式，而没有建立存储过程。开始的时候，他将几个表的访问代码放在了一起，导致WHAT参数输入的字符串超长而报错，这里的错误显然就是ORA-1461，于是第二次使用DBMS_JOB.SUBMIT过程时，虽然输入参数这次满足要求了，但是引发上上面的bug。

Oracle给出的解决方法也很简单，设置cursor_space_for_time初始化参数为TRUE，或者打3315960的补丁。