Linuxha.

net
Step-by-step Guide:
Oracle 10g Express Edition

Last updated: 12 October 2006 Author: Colleen Romero

Introduction

The purpose of this document is to suggest procedures for creating a clustered Oracle 10g
Express Edition configuration with Linuxha.net. These procedures were tested with the
following configuration.
• Fedora Core 5
• DRBD version 0.7.20
• Linuxha.net version 1.2
• Oracle 10g Express Edition version 10.2.0.1

The following conventions are used throughout this document:

# ls Single-line command entered as root user.
# useradd \ Multi-line command entered as root user.
> --home-dir /home2/xyz \
> --gid abc \
> xyz
$ ls Single-line command entered as non-root user
$ cut \ Multi-line command entered as non-root user.
> --delimiter=":" \
> --fields=1,5 \
/ etc/passwd
127.0.0.1 localhost Command output or file contents
192.168.1.32 fc5s1
192.168.1.33 fc5s2
192.168.100.32 fc5s1b
192.168.100.33 fc5s2b

User Account Creation

The Oracle processes will run with the oracle user credentials. The oracle user and dba
group must exist on both nodes and have the same user and groups ids.

All commands in this section are executed as root on both nodes, unless otherwise indicated.

1. Check whether the oracle user exists and that the user and group ids are the same on
both nodes.
# grep oracle /etc/passwd
If there is no result, continue to step 2 since the oracle user does not exist. Otherwise,
compare the user and group ids on both nodes. These are the third and fourth fields
respectively of the user account record as shown below:
oracle:x:501:501::/usr/lib/oracle/xe:/bin/bash
If the user and group ids are not the same on both nodes, proceed to step 6. Otherwise
go to the next section.

2. Identify next available group id

# expr `cut --delimiter=":" --fields=3 /etc/group | \
> grep --regexp="^[5-9][0-9][0-9]$" | sort --numeric-sort|tail -1` + 1

3. Create dba group and assign the larger of the group ids identified in step 2, e.g. 504.
# groupadd --gid 504 dba

Oracle 10g Express Edition Page 2 of 14

4. Identify next available user id
# expr `cut --delimiter=":" --fields=3 /etc/passwd | \
> grep --regexp="^[5-9][0-9][0-9]$" | sort --numeric-sort|tail -1` + 1

5. Create oracle user and assign the larger of the user ids identified in step 4, e.g. 504.
# useradd --uid 504 --gid dba --home-dir /usr/lib/oracle-xe -M oracle

6. If the dba group ids are different, change the group id on one of the nodes to be the
same as the other, ensure that the new group id is not already in use.
# grep 504 /etc/group # check that group id 504 is not in use
# groupmod -g 504 dba # change group id to 504

7. If the oracle user ids are different, change the user id on one of the nodes to be the
same as the other, ensure that the new user id is not already in use.
# grep 504 /etc/passwd # check that user id 504 is not in use
# usermod --uid 504 --gid dba oracle # change user id to 504

8. Change the user and group ownership of files owned by oracle.

The following commands will find all files whose user / group ownership is 501 and
change their user / group ownership to oracle / dba.
# find /usr -gid 501 -exec chgrp dba {} \;
# find /usr -uid 501 -exec chown oracle {} \;

Installation

Carry out these instructions on both nodes, all commands are executed as root.

1. Check whether Oracle 10g Express Edition is installed, by executing:

# rpm --query --all|grep oracle
If the following is displayed then Oracle 10g Express Edition is installed and you may
proceed to step 4, otherwise continue to step 2.
oracle-xe-10.2.0.1-1.0

2. Download Oracle 10g Express Edition i386 RPM package (this is a single-line command).
# wget http://download.oracle.com/otn/linux/oracle10g/xe/10201/oracle-xe-
10.2.0.1-1.0.i386.rpm

3. Install Oracle 10g Express Edition

Execute the following:
# rpm --install --hash oracle-xe-10.2.0.1-1.0.i386.rpm
which should generate the following output
########################################### [100%]
########################################### [100%]
Executing Post-install steps...

You must run '/etc/init.d/oracle-xe configure' as the root user to

configure the database.

4. Configure Oracle 10g Express Edition

Oracle 10g Express Edition Page 3 of 14

 If Oracle 10g Express Edition was just installed then configure it by executing
# /etc/init.d/oracle-xe configure
accept the default ports, provide a password for the SYS and SYSTEM, enter n when
prompted whether to start Oracle 10g Express Edition at boot.
OR, if this is an existing Oracle 10g Express Edition installation that is configured to start
at boot then disable automatic startup by executing:
# /etc/init.d/oracle-xe disable

5. Edit /etc/init.d/oracle-xe
By default, the oracle-xe script will start or stop the application only if it is enabled at
boot. This behaviour has to be changed in order to use this script to start and stop the
application from the cluster.
Replace lines 597-628 in /etc/init.d/oracle-xe with the following:
case "$1" in
start)
start
;;
configure)
configure
;;
stop)
stop
;;

6. Start the Oracle database

# /etc/init.d/oracle-xe start

7. Delete the automatic startup and shutdown scripts

# rm /etc/rc0.d/K??oracle-xe /etc/rc3.d/S??oracle-xe

8. Configure firewall
If the firewall is running, it must be configured to permit access to the listener and HTTP
server ports defined in step 4. By default, these are ports 1521 and 8080 respectively.
Insert the following before the COMMIT command in /etc/sysconfig/iptables.
-A RH-Firewall-1-INPUT -m state --state NEW -m tcp -p tcp --dport 8080 -j ACCEPT
-A RH-Firewall-1-INPUT -m state --state NEW -m tcp -p tcp --dport 1521 -j ACCEPT

Restart the firewall, by executing:

# /etc/init.d/iptables restart

Directories/Files for Replication

All paths are relative to /usr/lib/oracle/xe

Directory/File Comment
oradata Database file directory
app/oracle/admin Database parameter file and alert log directory
app/oracle/flash_recovery_area Redo log, archive log, and RMAN backup
directory.

Oracle 10g Express Edition Page 4 of 14

 Directory/File Comment
app/oracle/product/10.2.0/server/dbs/spfileXE.ora Server parameter file for XE database
app/oracle/product/10.2.0/server/dbs/orapwXE File that stores names and passwords of XE
database users with the SYSDBA privilege
app/oracle/product/10.2.0/server/network/admin Network configuration script directory

Configuration

Steps 2 to 15 are performed as oracle, the rest as root.

Primary Node

1. Login as oracle
# su - oracle

2. Set the Oracle environment variables.

$ . $HOME/app/oracle/product/10.2.0/server/bin/oracle_env.sh
Append the command to ~/.bash_profile so that the environment variables are set
automatically.

3. Connect to the database as SYSDBA.

$ sqlplus "/ as sysdba"

4. Change the address of the default local listener to the virtual IP address (192.168.1.39).
SQL> ALTER SYSTEM
2 SET LOCAL_LISTENER='(ADDRESS = (PROTOCOL = TCP)
3 (HOST = 192.168.1.39)(PORT = 1521))';
SQL> ALTER SYSTEM REGISTER;

5. (Optional) Allow remote access to the database server.

SQL> EXEC DBMS_XDB.SETLISTENERLOCALACCESS(FALSE);

6. (Optional) Enable the HR account and assign a password.

SQL> ALTER USER HR IDENTIFIED BY HR ACCOUNT UNLOCK;

7. Exit SQL*Plus
SQL> EXIT

8. Create directory ~/app/oracle/admin/network for storage of listener configuration

scripts.
$ mkdir ~/app/oracle/admin/network

9. Copy the contents of $ORACLE_HOME/network/admin to ~/app/oracle/admin

/network
$ cp -r $ORACLE_HOME/network/admin/* ~/app/oracle/admin/network

10. Configure the Oracle listener to use the virtual IP address assigned to the clustered
application.
Replace the host name in line 16 of ~/app/oracle/admin/network/listener.ora with

Oracle 10g Express Edition Page 5 of 14

 the virtual IP address (192.168.1.39) as shown:
(ADDRESS = (PROTOCOL = TCP)(HOST = 192.168.1.39)(PORT = 1521))

11. Configure the database alias to point to the virtual IP address.

Replace the host name in line 5 of ~/app/oracle/admin/network/tnsnames.ora
with the virtual IP address as shown:
(ADDRESS = (PROTOCOL = TCP)(HOST = 192.168.1.39)(PORT = 1521))

12. Copy the initialization parameter file - spfileXE.ora - and the password file - orapwXE -
from $ORACLE_HOME/dbs to ~/app/oracle/admin/XE/dbs.
$ cp $ORACLE_HOME/dbs/spfileXE.ora $ORACLE_HOME/dbs/orapwXE
> ~/app/oracle/admin/XE/dbs

13. Create $ORACLE_HOME/server/dbs/initXE.ora as follows:

SPFILE=/usr/lib/oracle/xe/app/oracle/admin/XE/dbs/spfileXE.ora

14. Create a symbolic link to ~/app/oracle/admin/XE/dbs/orapwXE.

$ rm $ORACLE_HOME/dbs/orapwXE
$ ln -s ~/app/oracle/admin/XE/dbs/orapwXE \
> ~/app/oracle/product/10.2.0/server/dbs/orapwXE

15. Logout the oracle user.

$ exit

16. Shut down Oracle 10g Express Edition

# /etc/init.d/oracle-xe stop

17. Append the following to etc/profile:

TNS_ADMIN=/usr/lib/oracle/xe/app/oracle/admin/network;export TNS_ADMIN

Secondary Node

Steps 2 to 5 are performed as oracle, the rest as root.

1. Login as oracle
# su - oracle

2. Append the following to ~/.bash_profile:

. $HOME/app/oracle/product/10.2.0/server/bin/oracle_env.sh

3. Create ~/app/oracle/product/10.2.0/server/dbs/initXE.ora as follows:

SPFILE=/usr/lib/oracle/xe/app/oracle/admin/XE/dbs/spfileXE.ora

4. Create a symbolic link to ~/app/oracle/admin/XE/dbs/orapwXE.

$ rm $ORACLE_HOME/dbs/orapwXE
$ ln -s ~/app/oracle/admin/XE/dbs/orapwXE \
> ~/app/oracle/product/10.2.0/server/dbs/orapwXE

5. Logout the oracle user.

$ exit

Oracle 10g Express Edition Page 6 of 14

6. Shut down Oracle 10g Express Edition
# /etc/init.d/oracle-xe stop

7. Append the following line to /etc/profile:

TNS_ADMIN=/usr/lib/oracle/xe/app/oracle/admin/network;export TNS_ADMIN

Create Replicated File Systems

All commands are to be executed as root. Steps 1 to 3 must be executed on both nodes, the
rest on the primary node.

1. Use fdisk to create a partition (/dev/sdb5) to be used by the application volume group.
This should be at least 2 GB.

2. Initialize partition
# pvcreate /dev/sdb5

3. Create application volume group (oraclevg)

# vgcreate oraclevg /dev/sdb5

4. Create logical volume (adminlv) to store parameter files and alert logs, this should be
large enough to cater for growth of the alert logs.
# lvcreate --size 64M --name adminlv oraclevg

5. Create logical volume (flashlv) for flash recovery area, this needs to be large enough to
hold the database backups and redo logs (active and archived).
# lvcreate --size 512M --name flashlv oraclevg

6. Create logical volume (oradatalv) to store database files, this should be at least 1 GB.
# lvcreate --size 1200M --name oradatalv oraclevg

7. Create file systems on adminlv, flashlv and oradatalv.

# mkfs -t ext3 /dev/oraclevg/adminlv
# mkfs -t ext3 /dev/oraclevg/flashlv
# mkfs -t ext3 /dev/oraclevg/oradatalv

Populate Replicated File Systems

These instructions are to be carried out on the primary node only.

1. Copy contents of /usr/lib/oracle/xe/app/oracle/admin to adminlv.

# mount -t ext3 /dev/oraclevg/adminlv /mnt
# cd /usr/lib/oracle/xe/app/oracle/admin
# find . -print | cpio --pass-through /mnt
# umount /mnt

2. Copy contents of /usr/lib/oracle/xe/app/oracle/flash_recovery_area to flashlv.

# mount -t ext3 /dev/oraclevg/flashlv /mnt
# cd /usr/lib/oracle/xe/app/oracle/flash_recovery_area
# find . -print | cpio --pass-through /mnt

Oracle 10g Express Edition Page 7 of 14

 # umount /mnt

3. Copy contents of /usr/lib/oracle/xe/oradata to oradatalv.

# mount -t ext3 /dev/oraclevg/adminlv /mnt
# cd /usr/lib/oracle/xe/oradata
# find . -print | cpio --pass-through /mnt
# umount /mnt

Build Application

These instructions are to be carried out on the primary node ONLY.

1. Create directory /etc/cluster/oracle-xe

# mkdir /etc/cluster/oracle-xe

2. Create /etc/cluster/oracle-xe/appconf.xml as shown:

<?xml version="1.0"?>
<appconf>
<global>
<version>0.1</version>
<name>oracle-xe</name>
<takeover>normal</takeover>
<syncrate>2000</syncrate>
<preferred_node>fc5s1</preferred_node>
</global>
<networks>
<network net="main"
ip="192.168.1.39" netmask="255.255.255.0"
/>
</networks>
<vg>
<name>oraclevg</name>
<type>filesystems</type>
</vg>
<application>
<startscript>/etc/init.d/oracle-xe start</startscript>
<stopscript>/etc/init.d/oracle-xe stop</stopscript>
<maxstoptime>30</maxstoptime>
<maxstarttime>60</maxstarttime>
</application>
</appconf>

The bolded values are configuration-specific, as described in the following table:

Entry Value
global/preferred_node Host name of one of the nodes in the cluster, or LEAST_CPU_LOAD
networks/network.net Same as one of node/network.name in /etc/cluster/clconf.xml
networks/network.ip Virtual IP address of application
networks/network.netmask Netmask corresponding to virtual IP address
vg/name Name of application volume group

Oracle 10g Express Edition Page 8 of 14

3. Create /etc/cluster/oracle-xe/lems.local.xml as shown:
<?xml version="1.0"?>
<lems_config>
<globals
modules="/sbin/cluster/lems/modules"
programs="/sbin/cluster/lems/programs"
logs="/var/log/cluster/lems"
/>

<check>
<name>flag_check</name>
<type>internal</type>
<module>flag_check oracle-xe</module>
<interval>5</interval>
<action_list>
<action rc="0" action="NOP"/>
<action rc="1" action="%RCDATA%"/>
<action rc="2" action="ABORT"/>
</action_list>
</check>
<check>
<name>db</name>
<type>internal</type>
<module>procmon /etc/cluster/oracle-xe/db.xml</module>
<interval>10</interval>
<action_list>
<action rc="0" action="NOP"/>
<action rc="1" action="STOP"/>
<action rc="2" action="FAILOVER"/>
</action_list>
</check>
<check>
<name>listener</name>
<type>internal</type>
<module>procmon /etc/cluster/oracle-xe/listener.xml</module>
<interval>15</interval>
<action_list>
<action rc="0" action="NOP"/>
<action rc="1" action="STOP"/>
<action rc="2" action="FAILOVER"/>
</action_list>
</check>
<check>
<name>fsmonitor</name>
<type>internal</type>
<module>fsmon oracle-xe</module>
<interval>10</interval>
<action_list>
<action rc="0" action="NOP"/>
<action rc="1" action="PAUSE 30"/>
<action rc="2" action="STOP"/>
<action rc="3" action="FAILOVER"/>
<action rc="10" action="PAUSE 60"/>
</action_list>
</check>
</lems_config>

Oracle 10g Express Edition Page 9 of 14

4. Create /etc/cluster/oracle-xe/db.xml as shown:
<?xml version="1.0"?>
<procmon>
<global>
<logdir>/var/log/cluster</logdir>
<restarts>5</restarts>
<resetwindow>3600</resetwindow>
<restartcmd>/etc/init.d/oracle-xe restart </restartcmd>
</global>
<process>
<label>Database Writer Process</label>
<user>oracle</user>
<process_string>xe_dbw?_XE</process_string>
<min_count>1</min_count>
<max_count>10</max_count>
</process>
<process>
<label>Other Server Processes</label>
<user>oracle</user>
<process_string>xe_(lgwr|ckpt|pmon|reco|smon)_XE</process_string>
<min_count>5</min_count>
<max_count>5</max_count>
</process>
<!-- Uncomment this section if archive log is enabled -->
<!--
<process>
<label>Database Archiver Process</label>
<user>oracle</user>
<process_string>xe_arc?_XE</process_string>
<min_count>1</min_count>
<max_count>10</max_count>
</process>
-->
</procmon>

5. Create /etc/cluster/oracle-xe/listener.xml as shown:

<?xml version="1.0"?>
<procmon>
<global>
<logdir>/var/log/cluster</logdir>
<restarts>5</restarts>
<resetwindow>3600</resetwindow>
<restartcmd>
/bin/su - oracle -c "lsnrctl stop; lsnrctl start"
</restartcmd>
</global>
<process>
<label>Oracle 10g Express Edition Listener</label>
<user>oracle</user>
<process_string>
/usr/lib/oracle/xe/app/oracle/product/10.2.0/server/bin/tnslsnr*
</process_string>
<min_count>1</min_count>
<max_count>1</max_count>
</process>
</procmon>

Oracle 10g Express Edition Page 10 of 14

6. Verify LEMS configuration by executing:
# lems.pl --config /etc/cluster/oracle-xe/lems.local.xml \
> --application oracle-xe --verbose --check --file /dev/tty
If there are no errors, a result similar to the following will be returned:
INFO 09/09/2006 00:39:42 Using modules from : /sbin/cluster/lems/modules
INFO 09/09/2006 00:39:42 Using programs from : /sbin/cluster/lems/programs
INFO 09/09/2006 00:39:42 Writing logs to : /var/log/cluster/lems
INFO 09/09/2006 00:39:42 Listening on port : 9904
INFO 09/09/2006 00:39:42 Global initialisation complete.
INFO 09/09/2006 00:39:42 Started local server on port 9904
INFO 09/09/2006 00:39:42 Validating monitor entry fsmonitor...
INFO 09/09/2006 00:39:42 Validated monitor entry fsmonitor successfully.
INFO 09/09/2006 00:39:42 Validating monitor entry flag_check...
INFO 09/09/2006 00:39:42 Validated monitor entry flag_check successfully.
INFO 09/09/2006 00:39:42 Validating monitor entry oracle-xe...
INFO 09/09/2006 00:39:42 Validated monitor entry oracle-xe successfully.
INFO 09/09/2006 00:39:42 Check mode - transferring validated config to remote node.
INFO 09/09/2006 00:39:42 Configuration transferred successfully.
INFO 09/09/2006 00:39:42 Calculated a check interval of 2.5 seconds.

7. Mount the replicated file systems

# mount /dev/oraclevg/adminlv /usr/lib/oracle/xe/app/oracle/admin
# mount /dev/oraclevg/flashlv /usr/lib/oracle/xe/app/oracle/flash_recovery_area
# mount /dev/oraclevg/oradatalv /usr/lib/oracle/xe/oradata

8. Build oracle-xe.
# clbuildapp --application oracle-xe --sync
The following is the output of a successful build.
INFO 21/07/2006 23:37:06 Backups directory defaulted to /clbackup
INFO 21/07/2006 23:37:06
INFO 21/07/2006 23:37:06 Validation of Application 'oracle-xe' started.
INFO 21/07/2006 23:37:06 ['/var/log/cluster/build/oracle-xe-check-
300607212337.log']
INFO 21/07/2006 23:37:07 Initial Validation of Application successful.
INFO 21/07/2006 23:37:08
INFO 21/07/2006 23:37:08 NOTE: Build of new application is being performed.
INFO 21/07/2006 23:37:08
INFO 21/07/2006 23:37:08 Host Environment Validation started.
INFO 21/07/2006 23:37:08 ['/var/log/cluster/build/oracle-xe-envcheck-
300607212337.log']
INFO 21/07/2006 23:37:12 Host Environment Validation successful.
INFO 21/07/2006 23:37:12
INFO 21/07/2006 23:37:12 Cluster state : DOWN
INFO 21/07/2006 23:37:12 Application state: UNDEFINED
INFO 21/07/2006 23:37:12
INFO 21/07/2006 23:37:12 Volume Group Configuration started.
INFO 21/07/2006 23:37:12 ['/var/log/cluster/build/oracle-xe-lvm-
300607212337.log']
INFO 21/07/2006 23:37:20 Volume Group Configuration successful.
INFO 21/07/2006 23:37:20
INFO 21/07/2006 23:37:20 Application Resource Allocation started.
INFO 21/07/2006 23:37:20 ['/var/log/cluster/build/oracle-xe-build-
300607212337.log']
INFO 21/07/2006 23:37:33 Application Resource Allocation successful.
INFO 21/07/2006 23:37:33
INFO 21/07/2006 23:37:33 Application Data Synchronisation started.
INFO 21/07/2006 23:37:33 ['/var/log/cluster/build/oracle-xe-syncdata-
300607212337.log']
Storage Syncing: 1200Mb/ 1Mb [0.1 % Complete]
Storage Syncing: 0Mb/ 0Mb [100 % Complete]
INFO 21/07/2006 23:49:46 Application Data Synchronisation successful.
INFO 21/07/2006 23:49:47

Oracle 10g Express Edition Page 11 of 14

Run Application

For the commands used in this section, the host name of the primary node is fc5s1, and the
secondary node is fc5s2. All commands are executed as root on either node unless indicated
otherwise.

1. Form the cluster, if necessary.

# clform

2. Run oracle-xe on the primary node

# clrunapp --application oracle-xe --node fc5s1

3. Verify state of oracle-xe using clstat

# clstat --application oracle-xe
If oracle-xe is running, a result similar to the following will be displayed:
Cluster: cluster1 - UP

Application Node State Runnnig Monitor Stale Fail-over?

oracle-xe fc5s1 STARTED 0:00:00 Running 0 Yes

File Systems

Mount Point Valid Type State % Complete Completion

/usr/lib/oracle/xe/app/oracle/admin both drbd Sync
/usr/lib/oracle/xe/app/oracle/flash_recovery_area both drbd Sync
/usr/lib/oracle/xe/oradata both drbd Sync

Process Monitors

Name Status Restarts Current Reset at

oracle-xe Running 5 0 N/A

General Monitors

Type Name Status

Flag Check flag_check Running
FS Monitor fsmonitor Running

4. On the node where oracle-xe is running, list the Oracle processes.

# ps -u oracle -f|sort --key=8
The following process list should be displayed:
UID PID PPID C STIME TTY TIME CMD
oracle 10991 1 0 11:57 ? 00:00:00
/usr/lib/oracle/xe/app/oracle/product/10.2.0/server/bin/tnslsnr LISTENER -inherit
oracle 11012 1 0 11:57 ? 00:00:00 xe_cjq0_XE
oracle 11006 1 0 11:57 ? 00:00:00 xe_ckpt_XE
oracle 11019 1 0 11:57 ? 00:00:00 xe_d000_XE
oracle 11002 1 0 11:57 ? 00:00:00 xe_dbw0_XE
oracle 11004 1 0 11:57 ? 00:00:00 xe_lgwr_XE
oracle 11000 1 0 11:57 ? 00:00:00 xe_mman_XE
oracle 11016 1 0 11:57 ? 00:00:00 xe_mmnl_XE
oracle 11014 1 0 11:57 ? 00:00:01 xe_mmon_XE
oracle 10996 1 0 11:57 ? 00:00:00 xe_pmon_XE
oracle 10998 1 0 11:57 ? 00:00:00 xe_psp0_XE
oracle 11040 1 0 11:57 ? 00:00:00 xe_q000_XE
oracle 11042 1 0 11:57 ? 00:00:00 xe_q001_XE
oracle 11031 1 0 11:57 ? 00:00:00 xe_qmnc_XE
oracle 11010 1 0 11:57 ? 00:00:00 xe_reco_XE
oracle 11021 1 0 11:57 ? 00:00:00 xe_s000_XE

Oracle 10g Express Edition Page 12 of 14

 oracle 11023 1 0 11:57 ? 00:00:00 xe_s001_XE
oracle 11025 1 0 11:57 ? 00:00:00 xe_s002_XE
oracle 11027 1 0 11:57 ? 00:00:00 xe_s003_XE
oracle 11008 1 0 11:57 ? 00:00:00 xe_smon_XE

5. Check listener status on the node where oracle-xe is running.

$ su - oracle -c "lsnrctl stat"
The listener should return a status similar to the following:
Connecting to (DESCRIPTION=(ADDRESS=(PROTOCOL=IPC)(KEY=EXTPROC_FOR_XE)))
STATUS of the LISTENER
------------------------
Alias LISTENER
Version TNSLSNR for Linux: Version 10.2.0.1.0 - Production
Start Date 24-JUL-2006 12:05:19
Uptime 0 days 1 hr. 11 min. 37 sec
Trace Level off
Security ON: Local OS Authentication
SNMP OFF
Default Service XE
Listener Parameter File /usr/lib/oracle/xe/app/oracle/product/10.2.0/server
/network/admin/listener.ora
Listener Log File /usr/lib/oracle/xe/app/oracle/product/10.2.0/server
/network/log/listener.log
a
Listening Endpoints Summary...
(DESCRIPTION=(ADDRESS=(PROTOCOL=ipc)(KEY=EXTPROC_FOR_XE)))
(DESCRIPTION=(ADDRESS=(PROTOCOL=tcp)(HOST=192.168.1.39)(PORT=1521)))
(DESCRIPTION=(ADDRESS=(PROTOCOL=tcp)(HOST=fc5s1)(PORT=8080))(Presentation=HTTP)(
Session=RAW))
b Services Summary...
Service "PLSExtProc" has 1 instance(s).
Instance "PLSExtProc", status UNKNOWN, has 1 handler(s) for this service...
Service "XE" has 1 instance(s).

}
Instance "XE", status READY, has 1 handler(s) for this service...
Service "XEXDB" has 1 instance(s).
c
Instance "XE", status READY, has 1 handler(s) for this service...
Service "XE_XPT" has 1 instance(s).
Instance "XE", status READY, has 1 handler(s) for this service...
The command completed successfully

Referring to the bolded lines:

a) The listener should be running on the cluster application virtual IP address
"HOST=192.168.1.39"
b) The address for the HTTP service should be either the node name (HOST=fc5s1) or
the virtual IP address (HOST=192.168.1.39)
c) These six lines are displayed if the database has registered with the listener.

6. Verify database connection using SQL*Plus on either node, by executing the following:
# su - oracle -c "sqlplus -S /nolog" << EOF
> CONNECT HR/HR@XE
> COL REGION_NAME FOR A30
> SELECT * FROM REGIONS;
> EXIT
> EOF
The following output would be generated if the connection was successful:
REGION_ID REGION_NAME
---------- ------------------------------
1 Europe
2 Americas
3 Asia
4 Middle East and Africa

Oracle 10g Express Edition Page 13 of 14

7. Verify remote access via HTTP from any networked computer by connecting to the apex
application on the virtual IP address (192.168.1.39) at port 8080 from a web browser or
by executing:
# curl http://192.1.168.39:8080/apex
If the HTTP server is accepting connections on the virtual IP address, curl will produce no
output, otherwise an error message will be returned.

8. Stop oracle-xe by executing the following on the node where it is running:

# clhaltapp --application oracle-xe

9. Start oracle-xe on the secondary node (fc5s2)

# clrunapp --application oracle-xe --node fc5s2

10. Repeat steps 3-8.

Oracle 10g Express Edition Page 14 of 14