A Practical Guide to Oracle 10g RAC

Its REAL Easy!

Gavin Soorma,
Emirates Airline, Dubai
Session# 106
Agenda
• RAC concepts
• Planning for a RAC installation
• Pre Installation steps
• Installation of 10g R2 Clusterware
• Installation of 10g R2 Software
• Creation of RAC database
• Configuring Services and TAF
• Migration of single instance to RAC
What Is a RAC Cluster?

• Nodes
• Interconnect Interconnect

• Shared disk subsystem

• Node
Instances Node
• Database

Disks
Database vs Instance
RAC Cluster consists of ….
One or more instances
One Database residing on shared storage

Instance 1 Interconnect Instance 2

Node 1 Node 2

Local Shared Local

Disk Storage Disk

Database
 Why RAC?
• High Availability – survive node and instance failures

• Scalability – Add or remove nodes when needed

• Pay as you grow – harness the power of multiple low-cost

computers

• Enable Grid Computing

• DBA’s have their own vested interests!

 What is Real Application Clusters?

• Two or more interconnected, but independent servers

• One instance per node

• Multiple instances accessing the same database

• Database files stored on disks physically or logically connected to

each node, so that every instance can read from or write to them
A RAC Database –what’s different?
Contents similar to single instance database except …

•Create and enable one redo thread per instance

•If using Automatic Undo Management also require one UNDO tablespace
per instance

•Additional cluster specific data dictionary views created by

running the script $ORACLE_HOME/rdbms/admin/catclust.sql

•New background processes

•Cluster specific init.ora parameters

RAC specific Background Processes

• LMON – Global Enqueue Service Monitor

• LMD0 – Global Enqueue Service Daemon
• LMSx – Global Cache Server Processes
• LCK0 – Lock Process
• DIAG – Diagnosability Process
RAC init.ora Parameters
*.db_cache_size=113246208
*.java_pool_size=4194304
*.db_name='racdb‘
racdb2.instance_number=2
racdb1.instance_number=1
*.log_archive_dest_1='LOCATION=USE_DB_RECOVERY_FILE_DEST'
racdb2.thread=2
racdb1.thread=1
*.undo_management='AUTO'
racdb2.undo_tablespace='UNDOTBS2'
racdb1.undo_tablespace='UNDOTBS1'
 10g RAC Implementation Steps

• Hardware – Network Interface Cards, HBA cards etc

• Interconnects – Physical cable, Gigabit Ethernet switch
• Network – Virtual IP addresses
• Plan the type of shared storage (ASM, OCFS etc)
• Download latest RPM’s – ASM, OCFS
• Install Clusterware (Cluster Ready Services)
• Install 10g RAC software
• Create RAC database
• Configure Services and TAF ( Transparent Application Failover)
 RAC Database Storage
• Oracle files (control file, data files, redo log files)
• Server Parameter File ( SPFILE)
• Archive log files
• Flash Recovery Area
• Voting File
• Oracle Cluster Registry (OCR) File
• OCFS version 2.x will support shared ORACLE_HOME
 Oracle Cluster Registry File

• OCR contains important metadata about RAC instances and nodes

that make up the cluster

• Needs to be on a shared storage device

• About 100MB in size

• In Oracle 10g Release 2, higher availability for this critical

component is provided by enabling a second OCR file location
 Voting Disk File

• Contains information about cluster membership

• Used by CRS to avoid ‘split-brain’ scenarios if any node loses

contact over the interconnect

• Mandatory to be located on shared storage

• Typically about 20MB in size

• Can be mirrored in Oracle 10g Release 2

Shared Storage Considerations
Mandatory for:
Datafiles, Redo Log Files, Control Files, SPFile
Optional for:
Archive logfiles, Executables, Binaries, Network Configuration
files
Supported shared storage
NAS (network attached storage)
SAN ( storage area network)
Supported file storage
Raw volumes
Cluster File System
ASM
 Shared Storage Considerations
• Archive log files cannot be placed on raw devices

• CRS Files ( Voting Disk/Cluster Registry (OCR) cannot

be stored on ASM

• Software is installed on regular file system local to each node

• Database files can exist on raw devices, ASM or

Cluster File System (OCFS)
Network Requirements
• Each node must have at least two network adapters; one for the
public network interface and one for the private network
interface (the interconnect).

• The public network adapter must support TCP/IP

• For the private network, the interconnect must preferably be a

Gigabit Ethernet switch that supports UDP. This is used for
Cache Fusion inter-node connection

• Host name and IP addresses associated with the public

interface should be registered in DNS and /etc/hosts
IP Address Requirements

For each Public Network Interface an IP address and host

name registered in the DNS

One unused Virtual IP address and associated host name

registered in the DNS for each node to be used in the cluster

A private IP address and optional host name for each private

interface

Virtual IP addresses is used in the network config files

Virtual IP Addresses

• VIPs are used in order to facilitate faster failover in the event of a

node failure

• Each node not only has its own statically assigned IP address as well
as also a virtual IP address that is assigned to the node

• The listener on each node will be listening on the Virtual IP and client
connections will also come via this Virtual IP.

• Without VIP, clients will have to wait for long TCP/IP timeout before
getting an error message or TCP reset from nodes that have died
Sample /etc/hosts file

racdb1:/opt/oracle> cat /etc/hosts

# Do not remove the following line, or various programs

# that require network functionality will fail.
#127.0.0.1 itlinuxbl53.hq.emirates.com itlinuxbl53
localhost.localdomain localhost
57.12.70.59 itlinuxbl54.hq.emirates.com itlinuxbl54
57.12.70.58 itlinuxbl53.hq.emirates.com itlinuxbl53
10.20.176.74 itlinuxbl54-pvt.hq.emirates.com itlinuxbl54-pvt
10.20.176.73 itlinuxbl53-pvt.hq.emirates.com itlinuxbl53-pvt
57.12.70.80 itlinuxbl54-vip.hq.emirates.com itlinuxbl54-vip
57.12.70.79 itlinuxbl53-vip.hq.emirates.com itlinuxbl53-vip
Setup User equivalence using SSH

• To install on all nodes in the cluster by launching OUI on one node

• Will not prompt for password
• OUI will use ssh or rcp to copy files to remote nodes
ssh-keygen -t dsa
cat id_dsa.pub > authorized_keys
• Copy authorized_keys from this node to other nodes
• Run the same command on all nodes to generate the
authorized_keys file
• Finally all nodes will have the same authorized_keys file
 Setting up User Equivalence
ITLINUXBL53
ssh-keygen -t dsa
cat id_dsa.pub > authorized_keys
scp authorized_keys itlinuxbl54:/opt/oracle

ITLINUXBL54
ssh-keygen -t dsa
cat id_dsa.pub >> authorized_keys
scp authorized_keys itlinuxbl53:/opt/oracle/.ssh

ssh itlinuxbl54 hostname

ssh itlinuxbl53 hostname
Configure the hang check timer

• Monitors the Linux kernel for hangs

• If hang occurs module reboots the node
• Define how often in seconds module checks for hangs
• Define how long module waits for response from kernel

[root@itlinuxbl53 rootpre]# /sbin/insmod hangcheck-timer

hangcheck_tick=30 hangcheck_margin=180
Using /lib/modules/2.4.21-37.ELsmp/kernel/drivers/char/hangcheck-timer.o

[root@itlinuxbl53 rootpre]# lsmod | grep hang

hangcheck-timer 2672 0 (unused)
Case Study Environment
• Operating System: LINUX X86_64 RHEL 3AS
• Hardware: HP BL25P Blade Servers with 2 CPU’s (AMD 64 bit
processors) and 4 GB of RAM
• Oracle Software: Oracle Database 10g Enterprise Edition
Release 10.2.0.1.0 - 64bit
• Two Node Cluster: ITLINUXBL53.hq.emirates.com,
ITLINUXBL54.hq.emirates.com
• Shared Storage: OCFS for Cluster Registry and Voting Disks.
ASM for all other database related files
• Database Name: racdb
• Instance Names: racdb1, racdb2
Oracle 10g CRS Install

• Oracle 10g Clusterware – Cluster Ready Services

• Oracle’s own full-stack clusterware coupled with RAC
• Replaces earlier dependency on third-party clusterware
• Oracle CRS replaces the Oracle Cluster Manager (ORACM) in
Oracle9i RAC
• CRS must be installed prior to the installation of Oracle RAC
CRS Installation – Key Steps
• Voting Disk – about 20MB (Oracle9i Quorum Disk)
Maintains the node heartbeat and avoids the node split-brain
syndrome

• Oracle Cluster Registry – about 100MB

Stores cluster configuration and cluster database information

• Private Interconnect Information

Select the network interface for internode communication
A Gigabit Ethernet interface is recommended

• Run root.sh
Start CRS daemon processes – evmd, cssd, crsd
 Oracle Cluster File System
• Shared disk cluster file system for LINUX and Windows

• Improves management of data by eliminating the need to manage

raw devices

• Can be downloaded from OTN

http://oss.oracle.com/projects/ocfs

• OCFS 2.1.2 provides support on Linux for Oracle Software

installation as well
Install the OCFS RPM’s

[root@itlinuxbl54 recyclebin]# rpm -ivh ocfs-support-1.1.5-1.x86_64.rpm

Preparing... ########################################### [100%]
1:ocfs-support ########################################### [100%]

[root@itlinuxbl54 recyclebin]# rpm -ivh ocfs-tools-1.0.10-1.x86_64.rpm

Preparing... ########################################### [100%]
1:ocfs-tools ########################################### [100%]

[root@itlinuxbl54 recyclebin]# rpm -ivh ocfs-2.4.21-EL-smp-1.0.14-1.x86_64.rpm

Preparing... ########################################### [100%]
1:ocfs-2.4.21-EL-smp ########################################### [100%]
OCFSTOOL – Generate Config
 The OCFS Configuration File

[root@itlinuxbl53 etc]# cat /etc/ocfs.conf

#
# ocfs config
# Ensure this file exists in /etc
#
node_name = itlinuxbl53.hq.emirates.com
ip_address = 10.20.176.73
ip_port = 7000
comm_voting = 1
guid = 5D9FF90D969078C471310016353C6B23
OCFSTOOL – Format Partition
OCFSTOOL – Mount File System
OCFSTOOL – Mount File System
OCFSTOOL – Mount File System
 ASM Instance ASM Instance
RAC
Database
ASM Architecture Oracle Oracle
DB Instance DB Instance

Clustered
Servers

Disk Group Clustered

Pool of Storage
Install the ASMLIB RPM’s
[root@itlinuxbl53 recyclebin]# rpm -ivh oracleasm-support-2.0.1-1.x86_64.rpm
Preparing... ########################################### [100%]
1:oracleasm-support ###########################################
[100%]
[
[root@itlinuxbl53 recyclebin]# rpm -ivh oracleasm-2.4.21-37.ELsmp-1.0.4-
1.x86_64.rpm
Preparing... ########################################### [100%]
1:oracleasm-2.4.21-37.ELs###########################################

[root@itlinuxbl53 recyclebin]# rpm -ivh oracleasmlib-2.0.1-1.x86_64.rpm

Preparing... ########################################### [100%]
1:oracleasmlib ###########################################
[100%]
Creating the ASM Disks
[root@itlinuxbl53 init.d]# ./oracleasm createdisk VOL1 /dev/sddlmab1
Marking disk "/dev/sddlmab1" as an ASM disk: [ OK ]

[root@itlinuxbl53 init.d]# ./oracleasm createdisk VOL2 /dev/sddlmac1

Marking disk "/dev/sddlmac1" as an ASM disk: [ OK ]

[root@itlinuxbl53 init.d]# ./oracleasm createdisk VOL3 /dev/sddlmaf1

Marking disk "/dev/sddlmaf1" as an ASM disk: [ OK ]

[root@itlinuxbl53 init.d]# ./oracleasm listdisks

VOL1
VOL2
VOL3

[root@itlinuxbl54 init.d]# ./oracleasm scandisks

Scanning system for ASM disks: [ OK ]
The Cluster Verify Utility(cluvfy)

• Perform pre-installation and post-installation checks at

various stages of the RAC installation
• Available in 10g Release 2

./runcluvfy.sh comp nodereach -n itlinuxbl53,itlinuxbl54 –verbose

./runcluvfy.sh stage -pre crsinst -n itlinuxbl53,itlinuxbl54 –verbose
./runcluvfy.sh comp nodecon -n itlinuxbl53,itlinuxbl54 –verbose
./runcluvfy.sh stage -post hwos -n itlinuxbl53 -verbose
Install the cvuqdisk RPM for cluvfy
[root@itlinuxbl53 root]# cd /opt/oracle/cluster_cd/clusterware/rpm

[root@itlinuxbl53 rpm]# ls

cvuqdisk-1.0.1-1.rpm

[root@itlinuxbl53 rpm]# export CVUQDISK_GRP=dba

[root@itlinuxbl53 rpm]# rpm -ivh cvuqdisk-1.0.1-1.rpm

Preparing... ###########################################
[100%]
1:cvuqdisk ###########################################
[100%]
10g Clusterware Installation
Prerequisites Validation
Configuring the 10g RAC Cluster
Configuring the 10g RAC Cluster
Configuring the Network Interfaces
Oracle Cluster Registry (OCR)
Mirroring the OCR
Voting Disk
10g Clusterware OUI – Remote Installation
10g Clusterware – root.sh
Configuration Assistants
10g RAC phase one complete!
Verifying the Oracle Clusterware
Installation
Check node reachability
[oracle@itlinuxbl53 bin]$ ./olsnodes -n
itlinuxbl53 1
itlinuxbl54 2
Check for Clusterware processes
ps –ef |grep crs
ps –ef |grep css
ps –ef |grep evm

Check the health of the CRS stack

./crsctl check crs
10g RAC Software Installation
10g RAC Installation
10g RAC Software Installation
10g RAC Software Installation
10g RAC Software Installation
10g RAC Software Installation
Remote Node Installation
10g RAC Software Installation
Creating the RAC Database using DBCA
Configuring ASM
Creating the ASM Instances
Creating the ASM Instances
Creating the ASM Instances
Creating the ASM Instances
Creating the ASM Disk Groups
Creating the ASM Disk Groups
Creating the ASM Disk Groups
Creating the ASM Disk Groups
Creating the ASM Disk Groups
Creating the RAC Database using DBCA
DBCA is also Cluster aware
Creating the RAC Database using DBCA
Creating the RAC Database using DBCA
Creating the RAC Database using DBCA
Creating the RAC Database using DBCA
Creating the RAC Database using DBCA
Creating the RAC Database using DBCA
Creating the RAC Database using DBCA
Creating the RAC Database using DBCA
Enabling Flashback & Archive
logging
•Archive log files preferably need to be located on
shared storage – in this case the ASM Disk Group

SQL> alter system set db_recovery_file_dest_size=2G scope=both

sid='*';
SQL> alter system set db_recovery_file_dest='+DG1' scope=both
sid='*';
SQL> alter system set
log_archive_dest_1='location=USE_DB_RECOVERY_FILE_DEST'
scope=both sid='*';
SQL> shutdown immediate;
Note: shutdown the other instances as well
Enabling Flashback & Archive
logging
• Connect to one of the instances in the RAC cluster and mount
the instance

SQL> startup mount;

SQL> alter database archivelog;
SQL> alter database open;

• Startup other instances in the RAC cluster as well

SQL> archive log list

Database log mode Archive Mode
Automatic archival Enabled
Archive destination USE_DB_RECOVERY_FILE_DEST
Oldest online log sequence 15
Next log sequence to archive 16
Current log sequence 16
 Services

• Logically group consumers who share common attributes like

workload, a database schema or some common application
functionality
• Manage client load balancing
• Manage server-side load balancing
• Connect-time failover with TAF
• Controlled by tnsnames.ora parameters FAILOVER=ON,
FAILOVER_MODE, METHOD
• Managed via DBCA or SRVCTL commands
Configuring Services
Configuring Services
Configuring Services
Configuring Services
Configuring Services
Managing services (srvctl)
racdb2:/opt/oracle/product/10.2.0/db/bin>srvctl status asm -n itlinuxbl53
ASM instance +ASM1 is running on node itlinuxbl53.

racdb2:/opt/oracle/product/10.2.0/db/bin>srvctl config database -d racdb

itlinuxbl53 racdb1 /opt/oracle/product/10.2.0/db
itlinuxbl54 racdb2 /opt/oracle/product/10.2.0/db

racdb2:/var/opt/oracle>srvctl start database -d racdb

racdb2:/var/opt/oracle>srvctl status database -d racdb
Instance racdb1 is running on node itlinuxbl53
Instance racdb2 is running on node itlinuxbl54

racdb2:/var/opt/oracle>srvctl config service -d racdb

racdb_blade53 PREF: racdb1 AVAIL: racdb2
racdb_blade54 PREF: racdb2 AVAIL: racdb1

racdb2:/var/opt/oracle>srvctl status service -d racdb -s racdb_blade53

Service racdb_blade53 is running on instance(s) racdb1
 Transparent Application Failover (TAF)

• TAF defined by FAILOVER_MODE parameter

• TYPE=SESSION
- User does not need to reconnect
- Session failed over to another available instance in the list
- But SQL statements in progress will have to be
reissued
• TYPE=SELECT
- Query will be restarted after failover
- Rows not fetched before failover will be retrieved
 Transparent Application Failover (TAF)
Connection modes METHOD=BASIC or PRECONNECT

• BASIC
- After failover connection must reconnect to next address in the list
- Additional time to failover

• PRECONNECT
- Session is opened against all addresses in the list
- Only one is used – others remain connected
- Faster failover with preconnected sessions
- More memory resources consumed by preconnected sessions on
other nodes
tnsnames.ora for RAC
Client-side load balancing

ERP =
(DESCRIPTION =
(LOAD_BALANCE = ON)
(FAILOVER=ON)
(ADDRESS_LIST =
(ADDRESS = (PROTOCOL = TCP)(HOST = rac1vip)(PORT = 1521))
(ADDRESS = (PROTOCOL = TCP)(HOST = rac2vip)(PORT = 1521)) (CONNECT_DATA =
(SERVICE_NAME = ERP.WORLD) (FAILOVER_MODE=(TYPE=SELECT)(METHOD=BASIC))

Server-side load balancing

*.REMOTE_LISTENERS=RACDB_LISTENERS – init.ora parameter

RACDB_LISTENERS=
(DESCRIPTION=
(ADDRESS=(PROTOCOL = TCP)(HOST = rac1vip)(PORT = 1521))
(ADDRESS = (PROTOCOL = TCP)(HOST =rac2vip)(PORT = 1521))
 Recovery and RAC
SQL> INSERT INTO SH.MYOBJECTS SELECT * FROM DBA_OBJECTS ;

SQL> DELETE FROM SH.MYOBJECTS;

SQL> COMMIT;

Now get the log sequence ….

SQL> SELECT SEQUENCE#, THREAD#, STATUS FROM V$LOG;

SEQUENCE# THREAD# STATUS

---------- ---------- ----------------
9 1 INACTIVE
10 1 CURRENT
4 2 ACTIVE
5 2 CURRENT
Recovery and RAC (contd)
RMAN> LIST BACKUP OF DATABASE SUMMARY;

List of Backups
===============
Key TY LV S Device Type Completion Time #Pieces #Copies Compressed Tag
------- -- -- - ----------- --------------- ------- ------- ---------- ---
12 B F A DISK 05-FEB-06 1 1 NO BACKUP_RACDB.HQ.EM_020506082809

Connect to instance racdb2 ……

RMAN> LIST BACKUP OF DATABASE SUMMARY;

List of Backups
===============
Key TY LV S Device Type Completion Time #Pieces #Copies Compressed Tag
------- -- -- - ----------- --------------- ------- ------- ---------- ---
12 B F A DISK 05-FEB-06 1 1 NO BACKUP_RACDB.HQ.EM_020506082809

Since the same control files are used by both instances RACDB1 and RACDB2
(same db RACDB) the output is the same on both sides.
Recovery and RAC (contd)
srvctl stop database -d RACDB
srvctl start database -d RACDB -o mount

export ORACLE_SID=racdb1

run {
set until logseq 10 thread 1;
set autolocate on;
allocate channel c1 type disk;
restore database ;
recover database ;
release channel c1;
}
 Recovery and RAC (contd)
Starting restore at 05-FEB-06

channel c1: starting datafile backupset restore

channel c1: specifying datafile(s) to restore from backup set
restoring datafile 00001 to +DG1/racdb/datafile/system.256.1
……..
……..
piece handle=+DG1/racdb/backupset/2006_02_05/nnndf0_tag20060205t101703_0.665.7 tag=TAG20060205T101703
channel c1: restore complete
Finished restore at 05-FEB-06

Starting recover at 05-FEB-06

starting media recovery

archive log thread 1 sequence 9 is already on disk as file +DG1/racdb/archivelog/2006_02_05/thread_1_seq_9.653.7

archive log thread 2 sequence 4 is already on disk as file +DG1/racdb/archivelog/2006_02_05/thread_2_seq_4.662.7

RMAN> sql ' alter database open resetlogs';

 Migrate a Single-instance database to
RAC

• Create the directory structure for the database files and archive
log files on the OCFS file system

$ cd /ocfs/oradata/
$ mkdir gavin
$ cd /ocfs/oradata/gavin
$ mkdir arch
 Migrate a Single-instance database to
RAC

• Backup the current control file to trace

SQL> alter database backup controlfile to trace;

• Edit the CREATE CONTROLFILE script to change the location

of all the datafiles and redo log files to the OCFS file system
 Migrate a Single-instance database to
RAC
CREATE CONTROLFILE REUSE DATABASE "GAVIN" NORESETLOGS ARCHIVELOG
MAXLOGFILES 16
MAXLOGMEMBERS 3
MAXDATAFILES 100
MAXINSTANCES 8
MAXLOGHISTORY 454
LOGFILE
GROUP 1 '/ocfs/oradata/gavin/redo01.log' SIZE 10M,
GROUP 2 '/ocfs/oradata/gavin/redo02.log' SIZE 10M,
GROUP 3 '/ocfs/oradata/gavin/redo03.log' SIZE 10M
-- STANDBY LOGFILE
DATAFILE
'/ocfs/oradata/gavin/system01.dbf',
'/ocfs/oradata/gavin/undotbs01.dbf',
'/ocfs/oradata/gavin/sysaux01.dbf',
'/ocfs/oradata/gavin/users01.dbf',
'/ocfs/oradata/gavin/example01.dbf'
CHARACTER SET WE8ISO8859P1
;
 Migrate a Single-instance database to
RAC
• Shutdown the database and copy the files from the original
location to the OCFS location

$ cd /u01/ORACLE/gavin/

$ ls
arch control03.ctl redo02.log system01.dbf users01.dbf
control01.ctl example01.dbf redo03.log temp01.dbf
control02.ctl redo01.log sysaux01.dbf undotbs01.dbf

$ cp *.* /ocfs/oradata/gavin
 Migrate a Single-instance database to
RAC

• Change the location of the control files in the init.ora

*.control_files='/ocfs/oradata/gavin/control01.ctl','/ocfs/oradata/ga
vin/control02.ctl','/ocfs/oradata/gavin/control03.ctl‘

• Run the script to recreate the controlfile

SQL> startup nomount;

SQL> @crectl
 Migrate a Single-instance database to
RAC
•Note that the new location of the datafiles of the database is now
the shared OCFS file system

SQL> select name from v$datafile;

NAME
--------------------------------------------------------------------------------
/ocfs/oradata/gavin/system01.dbf
/ocfs/oradata/gavin/undotbs01.dbf
/ocfs/oradata/gavin/sysaux01.dbf
/ocfs/oradata/gavin/users01.dbf
/ocfs/oradata/gavin/example01.dbf
 Migrate a Single-instance database to
RAC

• Create the cluster specific data dictionary views by running

the catclust.sql script

SQL> @?/rdbms/admin/catclust.sql
Migrate a Single-instance database to
RAC
• Each instance in the cluster needs to have access to its own
thread of online redo log files.

• Create another thread of online log files

SQL> alter database add logfile thread 2

2 group 4 ('/ocfs/oradata/gavin/redo04.log') size 10m,
3 group 5 ('/ocfs/oradata/gavin/redo05.log') size 10m,
4 group 6 ('/ocfs/oradata/gavin/redo06.log') size 10m;
Database altered.

• Enable the thread

SQL> alter database enable public thread 2;

Database altered.
 Migrate a Single-instance database to
RAC

• Each instance needs to have its own dedicated undo

tablespace as well

SQL> create undo tablespace undotbs2 datafile

2 '/ocfs/oradata/gavin/undotbs02.dbf' size 200m;
Tablespace created
Migrate a Single-instance database to
RAC
• Make the following changes to the Init.ora parameter file:

ADD the following entries

*.cluster_database=TRUE
*.cluster_database_instances=2
gavin1.instance_name=gavin1
gavin2.instance_name=gavin2
gavin1.instance_number=1
gavin2.instance_number=2
gavin1.thread=1
gavin2.thread=2
gavin1.undo_tablespace=UNDOTBS1
gavin2.undo_tablespace=UNDOTBS2
*.remote_listener='LISTENERS_GAVIN'
 Migrate a Single-instance database to
RAC
Make the following changes to the Init.ora parameter file:

EDIT the following entries

gavin.__db_cache_size=171966464
gavin.__java_pool_size=8388608
gavin.__large_pool_size=4194304
gavin.__shared_pool_size=75497472

Change to ...

*.__db_cache_size=171966464
*.__java_pool_size=8388608
*.__large_pool_size=4194304
*.__shared_pool_size=75497472
 Migrate a Single-instance database to
RAC

• Change the archive log destination to the shared

disk as all instances need access to the archive log
files generated by each individual instance.

*.log_archive_dest_1='LOCATION=/ocfs/oradata/gavin/arch/‘

• Create the password file on each node

$ cd $ORACLE_HOME/dbs
$ orapwd file=orapwgavin1 password=oracle
 Migrate a Single-instance database to
RAC

• Add the following lines to the tnsnames.ora file on

BOTH NODES

LISTENERS_GAVIN =
(ADDRESS_LIST =
(ADDRESS = (PROTOCOL = TCP)(HOST =
hqlinuxrac101.hq.emirates.com)(PORT = 1521))
(ADDRESS = (PROTOCOL = TCP)(HOST =
hqlinuxrac102.hq.emirates.com)(PORT = 1521))
)
 Migrate a Single-instance database to
RAC
• Create the spfile which will be used by both
instances on the shared disk storage as well

SQL> create spfile='/ocfs/oradata/gavin/spfilegavin.ora' from

2 pfile='initgavin.ora';

• Create the init.ora for the instance gavin1 – only add

one line with the SPFILE value pointing to the spfile
we created on the OCFS file system
$ cat initgavin1.ora
SPFILE=/ocfs/oradata/gavin/spfilegavin.ora
Note: Do the same on the other node for the instance gavin2
 Migrate a Single-instance database to
RAC
• Start the instance on both nodes – first on Hqlinux05 and then
on Hqlinux06

SQL> select name from v$datafile;

NAME
--------------------------------------------------------------------------------
/ocfs/oradata/gavin/system01.dbf
/ocfs/oradata/gavin/undotbs01.dbf
/ocfs/oradata/gavin/sysaux01.dbf
/ocfs/oradata/gavin/users01.dbf
/ocfs/oradata/gavin/example01.dbf
/ocfs/oradata/gavin/undotbs02.dbf

SQL> select host_name from v$instance;

HOST_NAME
----------------------------------------------------------------
hqlinux06.hq.emirates.com
 Migrate a Single-instance database to
RAC
• Using the SRVCTL commands to configure services

$ srvctl add database -d gavin -o /opt/oracle/product10g/10.1.0.3

$ srvctl add instance -d gavin -i gavin1 -n hqlinux05
$ srvctl add instance -d gavin -i gavin2 -n hqlinux06
$ srvctl status instance -d gavin -i gavin1
Instance gavin1 is running on node hqlinux05
$ srvctl status instance -d gavin -i gavin2
Instance gavin2 is running on node hqlinux06
 Q&
Thanks for attending!!
GAVIN SOORMA
Technical Team Manager,
Databases
Emirates Airline, Dubai
Contact me at : 971507843900
or

A
gavin.soorma@emirates.com

QUESTIONS
ANSWERS
Contact me:
Email: gavin.soorma@emirates.com
Phone: + 971507843900
Acknowledgements & Thanks
• 10g RAC – Madhu Tumma
• High Availability with RAC, Flashback and Data
Guard – Matthew Hart & Scott Jesse
• A Rough Guide to RAC – Julian Dyke
• Oracle 10g Linux Administration – Edward Whalen