Bug 4369756 - Log apply may dump [krvsmso]

最新推荐文章于 2025-11-30 19:25:58 发布

转载最新推荐文章于 2025-11-30 19:25:58 发布 · 234 阅读

文章标签：

#数据库

本文记录了Oracle数据库中出现的ORA-07445和ORA-12805错误及解决方案，错误由LSP2进程与SQL Apply进程之间的竞态条件引起，导致多个LOGSTDBY Apply进程停止。通过打补丁可以解决该问题。

Sat Dec 4 04:05:58 2010

Errors in file /u01/app/oracle/admin/xxaadb/bdump/xxaadb_p012_549.trc:

ORA-07445: exception encountered: core dump [krvsmso()+1379] [SIGSEGV] [Address not mapped to object] [0x000000004] [] []

Sat Dec 4 04:06:32 2010

Errors in file /u01/app/oracle/admin/xxaadb/bdump/xxaadb_lsp0_2327.trc:

ORA-12805: parallel query server died unexpectedly

LOGSTDBY Apply process P020 pid=97 OS id=23901 stopped

LOGSTDBY Apply process P013 pid=50 OS id=23887 stopped

LOGSTDBY Apply process P024 pid=105 OS id=23909 stopped

LOGSTDBY Apply process P026 pid=109 OS id=23913 stopped

LOGSTDBY Apply process P018 pid=90 OS id=23897 stopped

LOGSTDBY Apply process P023 pid=104 OS id=23907 stopped

LOGSTDBY Apply process P021 pid=99 OS id=23903 stopped

LOGSTDBY Apply process P017 pid=81 OS id=23895 stopped

LOGSTDBY Apply process P025 pid=107 OS id=23911 stopped

LOGSTDBY Apply process P019 pid=93 OS id=23899 stopped

Sat Dec 4 04:06:32 2010

TLCR process death detected. Shutting down TLCR

logminer process death detected, exiting logical standby

LOGSTDBY Analyzer process P010 pid=46 OS id=23881 stopped

LOGSTDBY Apply process P011 pid=47 OS id=23883 stopped

LOGSTDBY Apply process P022 pid=100 OS id=23905 stopped

LOGSTDBY Apply process P016 pid=77 OS id=23893 stopped

LOGSTDBY Apply process P015 pid=76 OS id=23891 stopped

LOGSTDBY Apply process P014 pid=51 OS id=23889 stopped

Sat Dec 4 04:20:05 2010

RFS LogMiner: Client enabled and ready for notification

Sat Dec 4 04:20:06 2010

Primary database is in MAXIMUM PERFORMANCE mode

Cause
The cause of this problem has been identified and verified in an unpublished Bug 4369756. It is caused by a race condition between the LSP2 process as it validates object metadata and the other SQL Apply processes.

我的理解

lsp进程和sql apply 进程同时更新对象表I_OBJ#（object_id=3）或者tab$(object_id=4)引起的

Solution
The condition is cleared with the automatic restart of SQL Apply. No other action is necessary.

This issue is fixed in Oracle11g Release 1 and the 10.2.0.4 patchset. Some one-off patches may be available for 10.1.0.5 and up. Check Metalink for patch availability: Patch 4369756.

Oracle 说是会自动的restart sql apply,其实有时候apply 根本就停不下来。

解决方法就是打补丁。

来自 “ ITPUB博客 ” ，链接：http://blog.itpub.net/8135069/viewspace-681002/，如需转载，请注明出处，否则将追究法律责任。

转载于:http://blog.itpub.net/8135069/viewspace-681002/