标签归档：ipc

很多时候,linux/unix中的ipc未正常释放,表现的现象是再次启动实例的时候表现出来内存不足等类似现象,但是今天在数据库启动过程中因为数据库后台报ora-600[6002]错误,导致数据库启动失败,但是后续的一些列奇怪现象让我吃惊:
1.启动数据库异常的session一直处于卡主状态,任何办法无法终止
2.开打新会话,登录数据库提示进行操作提示not logging,但是abort会出现和1中一样的hang住
数据库open过程卡主 会话无法退出,crtl+c依然无效

SQL> startup
ORACLE instance started.

Total System Global Area  175775744 bytes
Fixed Size                  1343668 bytes
Variable Size             117444428 bytes
Database Buffers           50331648 bytes
Redo Buffers                6656000 bytes
Database mounted.


^C

^C^C^C^C^C

alert报错

Tue Aug 27 20:33:57 2013
SMON: enabling cache recovery
Archived Log entry 29 added for thread 1 sequence 29 ID 0xfb920b77 dest 1:
ARC3: Archival started
ARC0: STARTING ARCH PROCESSES COMPLETE
Errors in file /u01/oracle/diag/rdbms/ora11g/ora11g/trace/ora11g_ora_4992.trc  (incident=201905):
ORA-00600: internal error code, arguments: [6002], [6], [6], [1], [0], [], [], [], [], [], [], []
Incident details in: /u01/oracle/diag/rdbms/ora11g/ora11g/incident/incdir_201905/ora11g_ora_4992_i201905.trc
Use ADRCI or Support Workbench to package the incident.
See Note 411.1 at My Oracle Support for error and packaging details.
Undo initialization errored: err:600 serial:0 start:1524754 end:1528104 diff:3350 (33 seconds)
Errors in file /u01/oracle/diag/rdbms/ora11g/ora11g/trace/ora11g_ora_4992.trc:
ORA-00600: internal error code, arguments: [6002], [6], [6], [1], [0], [], [], [], [], [], [], []
Errors in file /u01/oracle/diag/rdbms/ora11g/ora11g/trace/ora11g_ora_4992.trc:
ORA-00600: internal error code, arguments: [6002], [6], [6], [1], [0], [], [], [], [], [], [], []
Error 600 happened during db open, shutting down database
USER (ospid: 4992): terminating the instance due to error 600
Instance terminated by USER, pid = 4992
ORA-1092 signalled during: alter database Open...

新开启会话登录

[oracle@xifenfei ~]$ sqlplus / as sysdba

SQL*Plus: Release 11.2.0.3.0 Production on Tue Aug 27 20:27:38 2013

Copyright (c) 1982, 2011, Oracle.  All rights reserved.

Connected.
SQL> startup mount
ORA-01012: not logged on
SQL> conn / as sysdba
Connected to an idle instance.
SQL> startup mount
ORA-01012: not logged on

alert日志提示

Tue Aug 27 20:53:18 2013
ORA-1092 : opitsk aborting process
Tue Aug 27 20:53:44 2013
ORA-1092 : opitsk aborting process
Tue Aug 27 20:54:21 2013

shutdown abort 依然卡主

SQL> shutdown abort


^C


^C^C

查看ipc
根据alert日志提示,数据库已经异常关闭了，但是sqlplus中显示异常，怀疑是ipc未被正常释放

[oracle@xifenfei ~]$ ipcs -m

------ Shared Memory Segments --------
key        shmid      owner      perms      bytes      nattch     status      
0x00000000 2228224    oracle    640        8388608    1                       
0x00000000 2260993    oracle    640        167772160  1                       
0xee5c29c4 2293762    oracle    640        2097152    1                       

[oracle@xifenfei ~]$ ps -ef|grep pmon
oracle    4843  4785  0 20:29 pts/3    00:00:00 grep pmon 

释放ipc

[oracle@xifenfei ~]$ ipcrm -m 2260993
[oracle@xifenfei ~]$ ipcs -m

------ Shared Memory Segments --------
key        shmid      owner      perms      bytes      nattch     status      
0x00000000 2228224    oracle    640        8388608    1                       
0x00000000 2260993    oracle    640        167772160  1          dest         
0xee5c29c4 2293762    oracle    640        2097152    1                       

[oracle@xifenfei ~]$ ipcrm -m 2228224
[oracle@xifenfei ~]$ ipcs -m

------ Shared Memory Segments --------
key        shmid      owner      perms      bytes      nattch     status      
0x00000000 2228224    oracle    640        8388608    1          dest         
0x00000000 2260993    oracle    640        167772160  1          dest         
0xee5c29c4 2293762    oracle    640        2097152    1                       

[oracle@xifenfei ~]$ ipcrm -m 2293762
[oracle@xifenfei ~]$ ipcs -m

------ Shared Memory Segments --------
key        shmid      owner      perms      bytes      nattch     status      
0x00000000 2228224    oracle    640        8388608    1          dest         
0x00000000 2260993    oracle    640        167772160  1          dest         
0x00000000 2293762    oracle    640        2097152    1          dest     

重新启动数据库

[oracle@xifenfei ~]$ sqlplus / as sysdba

SQL*Plus: Release 11.2.0.3.0 Production on Tue Aug 27 20:33:35 2013

Copyright (c) 1982, 2011, Oracle.  All rights reserved.

Connected to an idle instance.

SQL> startup mount;
ORACLE instance started.

Total System Global Area  175775744 bytes
Fixed Size                  1343668 bytes
Variable Size             117444428 bytes
Database Buffers           50331648 bytes
Redo Buffers                6656000 bytes
Database mounted.

再次查看ipc

[oracle@xifenfei ~]$ ipcs -m

------ Shared Memory Segments --------
key        shmid      owner      perms      bytes      nattch     status      
0x00000000 2228224    oracle    640        8388608    1          dest         
0x00000000 2260993    oracle    640        167772160  1          dest         
0x00000000 2293762    oracle    640        2097152    1          dest         
0x00000000 2490374    oracle    640        8388608    29                      
0x00000000 2523143    oracle    640        167772160  29                      
0xee5c29c4 2555912    oracle    640        2097152    29