# 43419
Wangzwang
|
|
To: to eygle
在进行编译存储过程的时候Oracle.exe进程的cpu占用率99%,始终编译不过去,没有提示语法错误,编译就没有任何反应了, 过程也没有语法问题, 把存储过程改为函数 或者把它放到Packages中也是同样的问题,我的Oracle版本是Version 9.2.0.1.0 ,请问一下这个一般是什么问题,是不是Oracle的Bug, 多谢 !
From: Wangzwang 2007.07.03 18:51
|
|
|
|
To: Wangzwang
[COLOR=blue]
看看v$session_wait,估计是有一些竞争。
[/COLOR]
From: eygle 2007.07.04 05:58
|
|
|
|
# 43418
Chi
来自: Mexico
|
|
To: Eygle
For your discussion of Oracle10g启动两个监听进程的故障报告, I had that problem before and I followed the temporary bug fix as metalink mentioned.I believer it should be
SUBSCRIBE_FOR_NODE_DOWN_EVENT_<local_listener_name>=OFF
in the listener.ora file
even with this, However, I just had the second listener again. Maybe I should the the patch.
From: Chi 2007.07.03 16:06
|
|
|
|
# 43417
Chi
来自: Mexico
|
|
To: Eygle
In your article: Oracle诊断案例-SGA与Swap之二, you identified the high SGA size verse SWAP space caused the busy system, so dropping the SGA solved the problem.Was this a general statement? i.e does it apply to other platform (IBM AIX 5.3) and other version of Oracle (like my 10g R2)
From: Chi 2007.07.03 07:43
|
|
|
|
To: Chi
[COLOR=blue]
那个讲的是一些经验值里面,这些经验具有普遍适用性,但是不是严格的。
[/COLOR]
From: eygle 2007.07.04 05:56
|
|
|
|
# 43410
Chi
来自: Mexico
|
|
To: Eygle
Finally convinced my counterpart to set my 10gR2 with archivelog and also multiplex the redo log (2 member per group) and archive logs. Because my server is a IBM p550 with AIX5.3L (3.8 G memory, 4 hd), so I set its lgpg_regions with the lock_sga=true.IN the meantimes, based others paper in AIX using with Oracle, I also I modified two parameters maxuproc (sys0)= 512, maxservers (aio0) =300.But all those settings caused serious paging problem even with just select cases.I observed very busy on hard drive specially on hdisk0, the system disk,high % of read.So how can I ease the paging from the Oracle side? which parameter setting I can modify on Oracle? (restart the Oracle is faster than reboot the server). Kind of urgent!
From: Chi 2007.06.29 23:39
|
|
|
|
# 43407
Julia
|
|
To:
准备洗碗。。。。。。
From: Julia 2007.06.28 05:39
|
|
|
|
# 43403
rocket
|
|
To: eygle
请盖老师讲解以下约束延时中的关键字DEFERRED,DEFERABLE,INITIALLY DEFERRED的区别、使用位置和时机,我看的资料有的地方用DEFERRED,有的地方用DEFERRABLE搞的我很迷糊。
From: rocket 2007.06.27 20:07
|
|
|
|
# 43402
Chi
来自: Mexico
|
|
To: Eygle
I checked many posts of English sites, Since RMAN does not backup online redo log, So There is always data discrepancy between a recovered Db and the original one.The last parts of transactions still hold in non-archived redo log will not be recovered (supposedly the redo log file is existed).Is that true?Unless I used the Data Guard configuration.
From: Chi 2007.06.27 15:20
|
|
|
|
To: Chi
[COLOR=blue]这是肯定的,DG也只有在最大保护模式下能够达到实时同步,否则必然存在差异。[/COLOR]
From: eygle 2007.07.02 23:25
|
|
|
|
# 43401
mfm088
来自: 北京
|
|
To:
这两天有一个客户他的RAC装在SOLARIS10上.两个节点.
这阵子监听器经常出问题.PS一查,出现两个监听器进程,这个时候应用程序无法连接数据库了.只能杀死全部两个进程,然后重启进程才行!去坛子上发贴一问, wlidflower 说经常遇到。这事客户可不这么看啊,可能觉得ORACLE有问题呢?不知您遇到过这种问题吗?
From: mfm088 2007.06.27 07:48
|
|
|
|
To: mfm088
[COLOR=blue]
这是一个已知的Bug,通常在较高负载下触发。
彻底修正需要打Patch的。
[/COLOR]
From: eygle 2007.06.27 20:05
|
|
|
|
|