pat98's always & forever

redhat 4에서 runInstaller 실행시 아래와 같이 에러가 발생한다면...(32bit 의 경우)

Unable to load native library: /tmp/OraInstall2009-11-12_04-37-29PM/jre/lib/i386/libjava.so: symbol __libc_wait, version GLIBC_2.0 not defined in file libc.so.6 with link time reference

거지 같습니다만. 설치 전에 아래와 같은 패치를 적용 시켜주어야 한다는것.

p3006854_9204_LINUX.zip

그냥 압축풀고 root 유저로 해당 shell 실행 시키고 다시 runInstaller 실행하면 잘 뜸니다. redhat 3라고 나와있지만 4도 해당 된다는거. shell을 돌리면 해당 library 의 permission을 변경합니다.

참 거지같은 오라클..하지만 해결되었으니 끝. 고민 안 합니다.

#-------------------------------------------------------------------------
#  Patch for Bug 3006854
#-------------------------------------------------------------------------
#
#  DATE:  October 22, 2003, Updated May 10th 2004
#  ----------------------------------------------------------
#  Platform Patch for : LINUX Intel
#
#  Bugs Fixed by this patch:
#  -------------------------
#  3006854, INSTALLER INTEROPERABILITY PATCH FOR RHEL3.0
#
#  This patch must be installed on systems running Red Hat Enterprise Linux 3.0
#  before running the Oracle Universal Installer (for any Oracle product).
#
#  This patch is not specific to a particular Oracle version.
#
#  Patch Installation Instructions:
#  --------------------------------
#  [1] To apply the patch, unzip the PSE container file:
#        % unzip p3006854_9204_LINUX.zip
#
#  [2] Login as the root user,
#
#  [3] Set your current directory to the directory where the patch
#      is located, eg:
#
#        # cd 3006854
#
#  [4] Run the script supplied by this patch
#
#        # sh rhel3_pre_install.sh
#
#  The installation of the patch is now complete. You can now run
#  the Oracle Universal Installer as normal.
#

invalid-file

Linux 11g R2 설치시 간단한 것이지만 바보짓을 할수 가 있네요.

이전 버전들에는 압축되어 있는 화일을 풀면 자동으로 /Disk1, Disk2 이런식으로 알아서 디렉토리가 생성되었지만 11g 압축화일은 화일 2개가 같은 내용이므로 같은 디렉토리에 풀면 됨. (풀면 쭈욱 /database 밑에 생성)

다시 말해 덮어 씌어지는 것이 아니라 이어서 다음 소스가 압축이 풀림.

$ unzip linux.x64_11gR2_database_1of2.zip
$ unzip linux.x64_11gR2_database_2of2.zip

1번만 풀고 runInstaller 실행시키면

database/stage/Components/oracle.jdk/
경로에 소스가 없다면서 에러가 남. 당연한 얘기 2번 화일에 다음 화일들이 이어지므로..

during unzip you will see the following line on the screen:
creating: database/stage/Components/oracle.jdk/
creating: database/stage/Components/oracle.jdk/1.5.0.17.0/
creating: database/stage/Components/oracle.jdk/1.5.0.17.0/1/
creating: database/stage/Components/oracle.jdk/1.5.0.17.0/1/DataFiles/
inflating: database/stage/Components/oracle.jdk/1.5.0.17.0/1/DataFiles/all.jar

해봤어? 정주영 말처럼 ㅋㅋ

Symptoms

Running a PL/SQL it's possible to hit the internal error:
   ORA-00600: internal error code, arguments: [kgiinb_invalid_obj], [0x67F45C7D8], [0x68ED1C1E0],

The call stack looks like:
   dbgexProcessError dbgeExecuteForError dbgePostErrorKGE dbkePostKGE_kgsf kgeadse
   kgerinv_internal kgerinv kgeasnmierr kgiinb pfri7_inst_body_com mon pfri3_inst_body pfrrun plsql_run
   peicnt kkxexe opiexe

Cause

The cause of this problem has been identified and verified in an Unpublished Bug 7420394.

Solution

Unpublished Bug 7420394 is fixed in 11.2 release.

To avoid the issue:

1. check if it's available oneoff Patch 7420394 on your current release and for your platform.

Applies to:

Goal

Solution

The DB server software provides the two scripts to configure automatic DB startup/shutdown with the server machine. They are
$ORACLE_HOME/bin/dbstart
$ORACLE_HOME/bin/dbshut

We need to call these scripts from the unix start/shutdown scripts (rc0.d / rc1.d etc.)

Step - 1: Check the oratab file in /etc/oratab or /var/opt/oracle/oratab
This should have the entry for the DB we are dealing with, with a value Y, like:
$ORACLE_SID:$ORACLE_HOME:Y

Step - 2: Login to root.
Save the following file in /etc/init.d/ (say the file name is dbora). Please note that /etc/init.d is RedHat specific.

-rwxr-xr-x 1 root root 1412 Aug 27 19:14 dbora

Mention the correct ORA_OWNER and ORA_HOME in the dbora

------------------ Start dbora ---------------------------------

#! /bin/bash
#
# oracle Start/Stop the Databases...
#
# chkconfig: 2345 99 10
#
# processname: oracle
# config: /etc/oratab
# pidfile: /var/run/oracle.pid

# Source function library.
. /etc/init.d/functions

RETVAL=0
ORA_OWNER="oracle"
ORA_HOME="/oracle/product/10.0.1"

# See how we were called.

prog="oracle"

start() {
echo -n $"Starting $prog: "
su - $ORA_OWNER -c "$ORA_HOME/bin/dbstart"
su - $ORA_OWNER -c "$ORA_HOME/bin/lsnrctl start"
RETVAL=$?
echo
[ $RETVAL -eq 0 ] && rm -f /var/lock/subsys/dbora

return $RETVAL
}

stop() {
echo -n $"Stopping $prog: "
su - $ORA_OWNER -c "$ORA_HOME/bin/dbshut"
su - $ORA_OWNER -c "$ORA_HOME/bin/lsnrctl stop"
}

restart() {
stop
start
}

case "$1" in
start)
start
;;
stop)
stop
;;
restart)
restart
;;
*)
echo $"Usage: $0 {start|stop|restart}"
exit 1
esac

exit $?


------------------ End dbora ---------------------------------


Step - 3: From root run the following:

# cd /sbin
# chkconfig --add dbora

This will create a system service viz. dbora
# chkconfig --list

This also creates the following files:

/etc/rc2.d/S99dbora ( calls $ORACLE_HOME/bin/dbstart )
/etc/rc3.d/S99dbora ( calls $ORACLE_HOME/bin/dbstart )
/etc/rc4.d/S99dbora ( calls $ORACLE_HOME/bin/dbstart )
/etc/rc5.d/S99dbora ( calls $ORACLE_HOME/bin/dbstart )

/etc/rc0.d/K10dbora ( calls $ORACLE_HOME/bin/dbshut )
/etc/rc1.d/K10dbora ( calls $ORACLE_HOME/bin/dbshut )
/etc/rc6.d/K10dbora ( calls $ORACLE_HOME/bin/dbshut )


chkconfig refers the " # chkconfig: 2345 99 10 " from dbora.
This signifies that the service has start run level set to 2, 3, 4 and 5.Stop run level set to 0, 1 and 6. And the start priority should be 99 and stop priority be 10.

If the version is Red Hat 3 ES or more than the dbora file needs two lines of comments.
Each  service which should be manageable by chkconfig needs two or more commented lines
added to its init.d script. The second line contains  a  description  for  the service, 
and may be extended across multiple lines with backslash continuation.

For example,
# chkconfig: 2345 99 10
# description: Saves and restores system entropy pool for higher quality random number generation

Step - 4: Reboot the Linux box and check