OEM 的OMS出现大量COREDUMP

今天突然一个OEM服务器，出现大量CORE FILE不一会儿

MOUNTPOINT就接近100%来了，紧急手动删除了文件，

但又不断生成，差了一下原来是OMS的APACHE的一个LOG大小

超过了2G（32位操作系统）

<ORACLE_HOME>/Apache/Apache/logs/access_log

 

照理说是自动做ROTATION的，但OEM的一个BUG，没有

去做，METALINK上的解决方法如下。

1. Shutdown the HTTP Server

2. Define rotation in $ORACLE_HOME/sysman/config/httpd_em.conf
i.e. replace the following 2 lines
ErrorLog <ORACLE_HOME>/Apache/Apache/logs/error_log
TransferLog <ORACLE_HOME>/Apache/Apache/logs/access_log
with
ErrorLog "|/<ORACLE_HOME>/Apache/Apache/bin/rotatelogs <ORACLE_HOME>/Apache/Apache/logs/error_log 43200"
TransferLog "|/<ORACLE_HOME>/Apache/Apache/bin/rotatelogs <ORACLE_HOME>/Apache/Apache/logs/access_log 43200"
where <ORACLE_HOME> is the full path to your OMS install directory.

3. Define rotation in $ORACLE_HOME/sysman/config/httpd_em.conf.template.
i.e. Replace the following two lines
ErrorLog &ORACLE_HOME&/Apache/Apache/logs/error_log
TransferLog "&ORACLE_HOME&/Apache/Apache/logs/access_log"
with
ErrorLog "|&ORACLE_HOME&/Apache/Apache/bin/rotatelogs &ORACLE_HOME&/Apache/Apache/logs/error_log 43200"
TransferLog "|&ORACLE_HOME&/Apache/Apache/bin/rotatelogs &ORACLE_HOME&/Apache/Apache/logs/access_log 43200"

4. Delete the existing access_log and error_log

5. Restart the HTTP Server

解决问题不难，关键在于找到问题所在