ORA-00600: internal error code, arguments: [4137]/ [4136]

现象描述：

xx系统oracle数据库11.2.0.4 反复重启：

后台日志报错如下：

SMON: enabling cache recovery
Undo initialization finished serial:0 start:173950654 end:173950754 diff:100 (1 seconds)
                                                                                                                                                                                          75589,1       98%
SMON: enabling cache recovery
Undo initialization finished serial:0 start:173950654 end:173950754 diff:100 (1 seconds)
Verifying file header compatibility for 11g tablespace encryption..
Verifying 11g file header compatibility for tablespace encryption completed
SMON: enabling tx recovery
Database Characterset is WE8ISO8859P1
Errors in file /oracle/db/diag/rdbms/sjxthx/sjxthx/trace/sjxthx_smon_30481.trc  (incident=516251):
ORA-00600: internal error code, arguments: [4137], [14.5.6274], [0], [0], [], [], [], [], [], [], [], []
Incident details in: /oracle/db/diag/rdbms/sjxthx/sjxthx/incident/incdir_516251/sjxthx_smon_30481_i516251.trc
Use ADRCI or Support Workbench to package the incident.
See Note 411.1 at My Oracle Support for error and packaging details.
No Resource Manager plan active
ORACLE Instance sjxthx (pid = 13) - Error 600 encountered while recovering transaction (14, 5).
Errors in file /oracle/db/diag/rdbms/sjxthx/sjxthx/trace/sjxthx_smon_30481.trc:
ORA-00600: internal error code, arguments: [4137], [14.5.6274], [0], [0], [], [], [], [], [], [], [], []
Mon Nov 13 15:53:46 2017
Sweep [inc][516251]: completed
Mon Nov 13 15:53:47 2017
Dumping diagnostic data in directory=[cdmp_20171113155347], requested by (instance=1, osid=30481 (SMON)), summary=[incident=516251].
replication_dependency_tracking turned off (no async multimaster replication found)
Errors in file /oracle/db/diag/rdbms/sjxthx/sjxthx/trace/sjxthx_smon_30481.trc  (incident=516252):
ORA-00600: internal error code, arguments: [4137], [14.5.6274], [0], [0], [], [], [], [], [], [], [], []
Incident details in: /oracle/db/diag/rdbms/sjxthx/sjxthx/incident/incdir_516252/sjxthx_smon_30481_i516252.trc
Use ADRCI or Support Workbench to package the incident.
See Note 411.1 at My Oracle Support for error and packaging details.
Mon Nov 13 15:53:47 2017
Errors in file /oracle/db/diag/rdbms/sjxthx/sjxthx/trace/sjxthx_p000_30521.trc  (incident=516355):
ORA-00600: internal error code, arguments: [2023], [0], [0], [], [], [], [], [], [], [], [], []
Incident details in: /oracle/db/diag/rdbms/sjxthx/sjxthx/incident/incdir_516355/sjxthx_p000_30521_i516355.trc
Dumping diagnostic data in directory=[cdmp_20171113155348], requested by (instance=1, osid=30481 (SMON)), summary=[abnormal process termination].
Use ADRCI or Support Workbench to package the incident.
See Note 411.1 at My Oracle Support for error and packaging details.
ORACLE Instance sjxthx (pid = 13) - Error 600 encountered while recovering transaction (14, 5).
Errors in file /oracle/db/diag/rdbms/sjxthx/sjxthx/trace/sjxthx_smon_30481.trc  (incident=516253):
ORA-00600: internal error code, arguments: [2023], [0], [0], [], [], [], [], [], [], [], [], []
Incident details in: /oracle/db/diag/rdbms/sjxthx/sjxthx/incident/incdir_516253/sjxthx_smon_30481_i516253.trc
Mon Nov 13 15:53:48 2017
Sweep [inc][516252]: completed
Dumping diagnostic data in directory=[cdmp_20171113155349], requested by (instance=1, osid=30521 (P000)), summary=[incident=516355].
Use ADRCI or Support Workbench to package the incident.
See Note 411.1 at My Oracle Support for error and packaging details.
SMON: Parallel transaction recovery slave got internal error
SMON: Downgrading transaction recovery to serial
Errors in file /oracle/db/diag/rdbms/sjxthx/sjxthx/trace/sjxthx_smon_30481.trc  (incident=516254):
ORA-00600: internal error code, arguments: [4137], [14.5.6274], [0], [0], [], [], [], [], [], [], [], []
Incident details in: /oracle/db/diag/rdbms/sjxthx/sjxthx/incident/incdir_516254/sjxthx_smon_30481_i516254.trc
Use ADRCI or Support Workbench to package the incident.
See Note 411.1 at My Oracle Support for error and packaging details.
Starting background process QMNC
Mon Nov 13 15:53:49 2017
QMNC started with pid=47, OS id=30583
ORACLE Instance sjxthx (pid = 13) - Error 600 encountered while recovering transaction (14, 5).
Mon Nov 13 15:53:50 2017
Sweep [inc][516254]: completed
Dumping diagnostic data in directory=[cdmp_20171113155350], requested by (instance=1, osid=30481 (SMON)), summary=[incident=516253].
Errors in file /oracle/db/diag/rdbms/sjxthx/sjxthx/trace/sjxthx_smon_30481.trc  (incident=516255):
ORA-00600: internal error code, arguments: [4136], [15.14.3879], [0], [0], [], [], [], [], [], [], [], []
Incident details in: /oracle/db/diag/rdbms/sjxthx/sjxthx/incident/incdir_516255/sjxthx_smon_30481_i516255.trc
Use ADRCI or Support Workbench to package the incident.
See Note 411.1 at My Oracle Support for error and packaging details.
ORACLE Instance sjxthx (pid = 13) - Error 600 encountered while recovering transaction (15, 14).
Errors in file /oracle/db/diag/rdbms/sjxthx/sjxthx/trace/sjxthx_smon_30481.trc:
ORA-00600: internal error code, arguments: [4136], [15.14.3879], [0], [0], [], [], [], [], [], [], [], []
Mon Nov 13 15:53:51 2017
Sweep [inc][516255]: completed
Dumping diagnostic data in directory=[cdmp_20171113155351], requested by (instance=1, osid=30481 (SMON)), summary=[incident=516255].
Starting background process SMCO
Mon Nov 13 15:53:52 2017
SMCO started with pid=50, OS id=30595
Mon Nov 13 15:53:55 2017
Completed: ALTER DATABASE OPEN



Mon Nov 13 15:21:58 2017
minact-scn: got error during useg scan e:1555 usn:14
minact-scn: useg scan erroring out with error e:1555
Mon Nov 13 15:21:59 2017
Starting background process CJQ0
Mon Nov 13 15:21:59 2017
CJQ0 started with pid=51, OS id=29559
Mon Nov 13 15:22:39 2017
Sweep [inc][456341]: completed
Sweep [inc][456231]: completed
Sweep [inc2][456341]: completed
Sweep [inc2][456233]: completed
Sweep [inc2][456232]: completed
Sweep [inc2][456231]: completed
Sweep [inc2][456230]: completed
Sweep [inc2][456229]: completed
Mon Nov 13 15:24:39 2017
minact-scn: got error during useg scan e:1555 usn:14
minact-scn: useg scan erroring out with error e:1555
Mon Nov 13 15:27:04 2017
Errors in file /oracle/db/diag/rdbms/sjxthx/sjxthx/trace/sjxthx_p000_29603.trc  (incident=456301):
ORA-00600: internal error code, arguments: [2023], [0], [0], [], [], [], [], [], [], [], [], []
Incident details in: /oracle/db/diag/rdbms/sjxthx/sjxthx/incident/incdir_456301/sjxthx_p000_29603_i456301.trc
Use ADRCI or Support Workbench to package the incident.
See Note 411.1 at My Oracle Support for error and packaging details.
Mon Nov 13 15:27:05 2017
Errors in file /oracle/db/diag/rdbms/sjxthx/sjxthx/trace/sjxthx_smon_29423.trc  (incident=456234):
ORA-00600: internal error code, arguments: [2023], [0], [0], [], [], [], [], [], [], [], [], []
Incident details in: /oracle/db/diag/rdbms/sjxthx/sjxthx/incident/incdir_456234/sjxthx_smon_29423_i456234.trc
Mon Nov 13 15:27:06 2017
Dumping diagnostic data in directory=[cdmp_20171113152706], requested by (instance=1, osid=29603 (P000)), summary=[incident=456301].
Use ADRCI or Support Workbench to package the incident.
See Note 411.1 at My Oracle Support for error and packaging details.
SMON: Parallel transaction recovery slave got internal error
SMON: Downgrading transaction recovery to serial
Mon Nov 13 15:27:06 2017
Sweep [inc][456301]: completed
Sweep [inc][456234]: completed
Sweep [inc2][456301]: completed
Errors in file /oracle/db/diag/rdbms/sjxthx/sjxthx/trace/sjxthx_smon_29423.trc  (incident=456235):
ORA-00600: internal error code, arguments: [4136], [15.14.3879], [0], [0], [], [], [], [], [], [], [], []
Sweep [inc2][456234]: completed
Incident details in: /oracle/db/diag/rdbms/sjxthx/sjxthx/incident/incdir_456235/sjxthx_smon_29423_i456235.trc
Use ADRCI or Support Workbench to package the incident.
See Note 411.1 at My Oracle Support for error and packaging details.
Mon Nov 13 15:27:08 2017
PMON (ospid: 29398): terminating the instance due to error 474
Mon Nov 13 15:27:09 2017
ORA-1092 : opitsk aborting process
System state dump requested by (instance=1, osid=29398 (PMON)), summary=[abnormal instance termination].
System State dumped to trace file /oracle/db/diag/rdbms/sjxthx/sjxthx/trace/sjxthx_diag_29409_20171113152709.trc
Dumping diagnostic data in directory=[cdmp_20171113152709], requested by (instance=1, osid=29398 (PMON)), summary=[abnormal instance termination].
Instance terminated by PMON, pid = 29398

问题分析：

ORA-00600: internal error code, arguments: [4136]这类问题基本都是由于undo段的问题，由于存储或断电当机故障操事务，如需要recover。

处理思路：

备份参数文件： 

 SQL>create pfile =’/home/oracle/xt2.ora’ from spfile;

修改pfile文件

*.undo_management='MANUAL'

#*.undo_tablespace='UNDOTBS1'

*.undo_tablespace='SYSTEM'

rollback_segments='SYSTEM'

启动数据库

Startup pfile=’/home/oracle/xt2.ora’

查看rollback segment状态

Select tablespace_name,status,segment_id,segment_name from dba_rollback_segs;

可以看到15,14回退段需要recover，再次关闭数据库，在参数文件里添加：

_CORRUPTED_ROLLBACK_SEGMENTS =( SYSSMU15_2717915580$,_SYSSMU14_431226599$)

_offline_rollback_segments=(_SYSSMU15_2717915580$,_SYSSMU14_431226599$)

SQL>startup pfile=’/home/oracle/xt2.ora’

方法一：
如果只是某个undo段有问题，而不想删除原来的undo表空间，可以采用下面的方法
SQL> drop rollback segment  "_SYSSMU14_431226599$";
SQL> drop rollback segment  "_SYSSMU15_2717915580$";
SQL>Create spfile from pfile=’/home/oracle/xt2.ora’
SQL>Shutdown immediate;
SQL>Startup;

方法二：
删除原来的undo表空间，并在参数文件里指定新的表空间
SQL> drop tablespace undotbs1 including contents and datafiles;
SQL>create undotbs02 tablespace datafile size 10g;
修改参数文件：
*.undo_tablespace='UNDOTBS02'
*.undo_management='AUTO'
重启数据库
SQL>Startup pfile=’/home/oracle/xt2.ora’
SQL>Create spfile from pfile=’/home/oracle/xt2.ora’
SQL>Shutdown immediate;
SQL>Startup;