DBA手记:共享内存无法正常释放的处理
在数据库启动之后,需要从操作系统上分配共享内存和信号量(Semaphore)资源,而在某些情况下,数据库异常关闭后,这些资源有可能无法正常释放,则在下次启动时,数据库可能遭遇错误,无法正常启动。
在一个客户环境(操作系统为SUN Solaris平台)中出现了ORA-04031错误之后,使用了shutdown abort选项关闭了数据库:
Sun Jul 19 19:20:23 2009
Errors in file /ora9i/oracle/admin/ora9i/bdump/ora9i1_j000_14337.trc:
ORA-12012: error on auto execute of job 721
ORA-04031: unable to allocate 4088 bytes of shared memory ("shared pool","STANDARDSYS","PL/SQL MPCODE","BAMIMA: Bam Buffer")
Sun Jul 19 19:24:24 2009
Errors in file /ora9i/oracle/admin/ora9i/bdump/ora9i1_j000_17497.trc:
ORA-12012: error on auto execute of job 721
ORA-04031: unable to allocate 4088 bytes of shared memory ("shared pool","STANDARDSYS","PL/SQL MPCODE","BAMIMA: Bam Buffer")
Sun Jul 19 19:26:56 2009
Shutting down instance (abort)
License high water mark = 1012
Instance terminated by USER, pid = 19157
接下来的尝试启动失败,告警日志记录了如下所示的错误信息:
Sun Jul 19 19:35:27 2009
Errors in file /ora9i/oracle/admin/ora9i/udump/ora9i1_ora_25055.trc:
ORA-27154: post/wait create failed
ORA-27300: OS system dependent operation:semget failed with status: 28
ORA-27301: OS failure message: No space left on device
ORA-27302: failure occurred at: sskgpsemsper
注意以上信息中,提示失败的操作是semget,其实如果仔细阅读这个提示,就可以获知错误的真正原因,在操作系统可以通过man semget查看手册,了解出错的真实原因。以下输出告诉我们,semget的任务是获得信号量集(get set of semaphores),所以也就可以知道之前的提示No space left on device并不是指存储空间,而是信号量资源:
oracle@db-server # man semget
NAME
semget - get set of semaphores
SYNOPSIS
#include
#include
#include
int semget(key_t key, int nsems, int semflg);
DESCRIPTION
The semget() function returns the semaphore identifier associated with key.
知道了这个原因以后,就可以通过ipcs命令来找到数据库的信号量资源占用,该主机上有三个数据库,Oracle用户的信号量集就是没有正常释放的那个:
oracle@db-server # ipcs -sa
IPC status from as of Sun Jul 19 22:01:09 CST 2009
T ID KEY MODE OWNER GROUP CREATOR CGROUP NSEMS OTIME CTIME
Semaphores:
s 851977 0x462fcb40 --ra-r----- orabil9 dbabil9 orabil9 dbabil9 604 22:07:58 12:54:47
s 1769482 0x6b997db0 --ra-r----- oracle9 dba9 oracle9 dba9 2004 22:07:15 17:03:39
s 1245195 0x8ea0d4d4 --ra-ra---- oracle dba oracle dba 2004 no-entry 19:27:06
接下来就可以通过操作系统的ipcrm命令
相关新闻>>
- 发表评论
-
- 最新评论 更多>>