Build Your Own Oracle RAC Cluster on Oracle Enterprise Linux and iSCSI (Continued)

The information in this guide is not validated by Oracle, is not supported by Oracle, and should only be used at your own risk; it is for educational purposes only.

Perform the following tasks on both Oracle RAC nodes in the cluster!

In this section we will create the oracle UNIX user account, recommended O/S groups, and all required directories. The following O/S groups will be created:

We will be using the Oracle Cluster File System, Release 2 (OCFS2) to store the files required to be shared for the Oracle Clusterware software. When using OCFS2, the UID of the UNIX user "oracle" and GID of the UNIX group "oinstall" must be the same on both of the Oracle RAC nodes in the cluster. If either the UID or GID are different, the files on the OCFS2 file system will show up as "unowned" or may even be owned by a different user. For this article, I will use 501 for the "oracle" UID and 501 for the "oinstall" GID.

Note that members of the UNIX group oinstall are considered the "owners" of the Oracle software. Members of the dba group can administer Oracle databases, for example starting up and shutting down databases. Members of the optional group oper have a limited set of database administrative privileges such as managing and running backups. The default name for this group is oper. To use this group, choose the "Custom" installation type to install the Oracle database software. In this article, we are creating the oracle user account to have all responsibilities!

Create Group and User for Oracle

Lets start this section by creating the UNIX oinstall and dba group and oracle user account:

# groupadd -g 501 oinstall
# groupadd -g 502 dba
# groupadd -g 503 oper
# useradd -m -u 501 -g oinstall -G dba,oper -d /home/oracle -s /bin/bash -c "Oracle Software Owner" oracle

# id oracle
uid=501(oracle) gid=501(oinstall) groups=501(oinstall),502(dba),503(oper)

Set the password for the oracle account:

# passwd oracle
Changing password for user oracle.
New UNIX password: xxxxxxxxxxx
Retype new UNIX password: xxxxxxxxxxx
passwd: all authentication tokens updated successfully.

Verify That the User nobody Exists

Before installing the Oracle software, complete the following procedure to verify that the user nobody exists on the system:

1. To determine if the user exists, enter the following command:

   # id nobody
   uid=99(nobody) gid=99(nobody) groups=99(nobody)

   If this command displays information about the nobody user, then you do not have to create that user.

2. If the user nobody does not exist, then enter the following command to create it:

   # /usr/sbin/useradd nobody

3. Repeat this procedure on all the other Oracle RAC nodes in the cluster.

Create the Oracle Base Directory

The next step is to create a new directory that will be used to store the Oracle Database software. When configuring the oracle user's environment (later in this section) we will be assigning the location of this directory to the $ORACLE_BASE environment variable.

The following assumes that the directories are being created in the root file system. Please note that this is being done for the sake of simplicity and is not recommended as a general practice. Normally, these directories would be created on a separate file system.

After the directory is created, you must then specify the correct owner, group, and permissions for it. Perform the following on both Oracle RAC nodes:

# mkdir -p /u01/app/oracle
# chown -R oracle:oinstall /u01/app/oracle
# chmod -R 775 /u01/app/oracle

At the end of this procedure, you will have the following:

• /u01 owned by root.
• /u01/app owned by root.
• /u01/app/oracle owned by oracle:oinstall with 775 permissions. This ownership and permissions enables the OUI to create the oraInventory directory, in the path /u01/app/oracle/oraInventory.

Create the Oracle Clusterware Home Directory

Next, create a new directory that will be used to store the Oracle Clusterware software. When configuring the oracle user's environment (later in this section) we will be assigning the location of this directory to the $ORA_CRS_HOME environment variable.

As noted in the previous section, the following assumes that the directories are being created in the root file system. This is being done for the sake of simplicity and is not recommended as a general practice. Normally, these directories would be created on a separate file system.

After the directory is created, you must then specify the correct owner, group, and permissions for it. Perform the following on both Oracle RAC nodes:

# mkdir -p /u01/app/crs
# chown -R oracle:oinstall /u01/app/crs
# chmod -R 775 /u01/app/crs

At the end of this procedure, you will have the following:

• /u01 owned by root.
• /u01/app owned by root.
• /u01/app/crs owned by oracle:oinstall with 775 permissions. These permissions are required for Oracle Clusterware installation and are changed during the installation process.

Create Mount Point for OCFS2 / Clusterware

Let's now create the mount point for the Oracle Cluster File System, Release 2 (OCFS2) that will be used to store the two Oracle Clusterware shared files.

Perform the following on both Oracle RAC nodes:

# mkdir -p /u02
# chown -R oracle:oinstall /u02
# chmod -R 775 /u02

Create Login Script for oracle User Account

To ensure that the environment is setup correctly for the "oracle" UNIX userid on both Oracle RAC nodes, use the following .bash_profile:

Note: When you are setting the Oracle environment variables for each Oracle RAC node, ensure to assign each RAC node a unique Oracle SID! For this example, I used:

• linux1: ORACLE_SID=racdb1
• linux2: ORACLE_SID=racdb2

Login to each node as the oracle user account:

# su - oracle

# .bash_profile

# Get the aliases and functions
if [ -f ~/.bashrc ]; then
      . ~/.bashrc
fi

alias ls="ls -FA"

export JAVA_HOME=/usr/local/java

# User specific environment and startup programs
export ORACLE_BASE=/u01/app/oracle
export ORACLE_HOME=$ORACLE_BASE/product/10.2.0/db_1
export ORA_CRS_HOME=/u01/app/crs
export ORACLE_PATH=$ORACLE_BASE/common/oracle/sql:.:$ORACLE_HOME/rdbms/admin
export CV_JDKHOME=/usr/local/java

# Each RAC node must have a unique ORACLE_SID. (i.e. racdb1, racdb2,...)
export ORACLE_SID=racdb1

export PATH=.:${JAVA_HOME}/bin:${PATH}:$HOME/bin:$ORACLE_HOME/bin
export PATH=${PATH}:/usr/bin:/bin:/usr/bin/X11:/usr/local/bin
export PATH=${PATH}:$ORACLE_BASE/common/oracle/bin
export ORACLE_TERM=xterm
export TNS_ADMIN=$ORACLE_HOME/network/admin
export ORA_NLS10=$ORACLE_HOME/nls/data
export NLS_DATE_FORMAT="DD-MON-YYYY HH24:MI:SS"
export LD_LIBRARY_PATH=$ORACLE_HOME/lib
export LD_LIBRARY_PATH=${LD_LIBRARY_PATH}:$ORACLE_HOME/oracm/lib
export LD_LIBRARY_PATH=${LD_LIBRARY_PATH}:/lib:/usr/lib:/usr/local/lib
export CLASSPATH=$ORACLE_HOME/JRE
export CLASSPATH=${CLASSPATH}:$ORACLE_HOME/jlib
export CLASSPATH=${CLASSPATH}:$ORACLE_HOME/rdbms/jlib
export CLASSPATH=${CLASSPATH}:$ORACLE_HOME/network/jlib
export THREADS_FLAG=native
export TEMP=/tmp
export TMPDIR=/tmp

Perform the following configuration procedures on both Oracle RAC nodes in the cluster!

The kernel parameters and shell limits discussed in this section will need to be defined on both Oracle RAC nodes in the cluster every time the machine is booted. This section provides information about setting those kernel parameters required for Oracle. Instructions for placing them in a startup script (/etc/sysctl.conf) are included in Section 16 ("All Startup Commands for Both Oracle RAC Nodes").

Overview

This section focuses on configuring both Oracle RAC Linux servers - getting each one prepared for the Oracle RAC 10g installation. This includes verifying enough swap space, setting shared memory and semaphores, setting the maximum number of file handles, setting the IP local port range, setting shell limits for the oracle user, activating all kernel parameters for the system, and finally how to verify the correct date and time for both nodes in the cluster.

There are several different ways to configure (set) these parameters. For the purpose of this article, I will be making all changes permanent (through reboots) by placing all commands in the /etc/sysctl.conf file.

Swap Space Considerations

• Installing Oracle Database 10g Release 2 on RHEL/OEL 5 requires a minimum of 1024MB of memory. (Note: An inadequate amount of swap during the installation will cause the Oracle Universal Installer to either "hang" or "die")
• To check the amount of memory you have, type:

  # cat /proc/meminfo | grep MemTotal
  MemTotal: 2074084 kB

• To check the amount of swap you have allocated, type:

  # cat /proc/meminfo | grep SwapTotal
  SwapTotal: 4128760 kB

• If you have less than 2048MB of memory (between your RAM and SWAP), you can add temporary swap space by creating a temporary swap file. This way you do not have to use a raw device or even more drastic, rebuild your system.

  As root, make a file that will act as additional swap space, let's say about 500MB:
  
  # dd if=/dev/zero of=tempswap bs=1k count=500000

  Now we should change the file permissions:
  
  # chmod 600 tempswap

  Finally we format the "partition" as swap and add it to the swap space:
  
  # mke2fs tempswap
  # mkswap tempswap
  # swapon tempswap
  

Configuring Kernel Parameters

Oracle Database 10g Release 2 on RHEL/OEL 5 requires the kernel parameter settings shown below. The values given are minimums, so if your system uses a larger value, do not change it:

kernel.shmmax = 4294967295
kernel.shmall = 268435456
kernel.shmmni = 4096
kernel.sem = 250 32000 100 128
fs.file-max = 65536
net.ipv4.ip_local_port_range = 1024 65000
net.core.rmem_default=1048576
net.core.rmem_max=1048576
net.core.wmem_default=262144
net.core.wmem_max=262144

RHEL/OEL 5 already comes configured with default values defined for the following kernel parameters:

kernel.shmall
kernel.shmmax
net.core.rmem_default
net.core.rmem_max
net.core.wmem_default
net.core.wmem_max

Use the default values if they are the same or larger than the required values.

This article assumes a fresh new install of Oracle Enterprise Linux 5 and as such, many of the required kernel parameters are already set (see above). This being the case, you can simply copy / paste the following to both Oracle RAC nodes while logged in as root:

# cat >> /etc/sysctl.conf <<EOF
kernel.shmmni = 4096
kernel.sem = 250 32000 100 128
fs.file-max = 65536
net.ipv4.ip_local_port_range = 1024 65000
EOF

The above command persisted the required kernel parameters through reboots by inserting them in the /etc/sysctl.conf startup file. Linux allows modification of these kernel parameters to the current system while it is up and running, so there's no need to reboot the system after making kernel parameter changes. To activate the new kernel parameter values for the currently running system, run the following as root on both Oracle RAC nodes in the cluster:

# sysctl -p
net.ipv4.ip_forward = 0
net.ipv4.conf.default.rp_filter = 1
net.ipv4.conf.default.accept_source_route = 0
kernel.sysrq = 0
kernel.core_uses_pid = 1
net.ipv4.tcp_syncookies = 1
kernel.msgmnb = 65536
kernel.msgmax = 65536
kernel.shmmax = 4294967295
kernel.shmall = 268435456
net.core.rmem_default = 1048576
net.core.rmem_max = 1048576
net.core.wmem_default = 262144
net.core.wmem_max = 262144
kernel.shmmni = 4096
kernel.sem = 250 32000 100 128
fs.file-max = 65536
net.ipv4.ip_local_port_range = 1024 65000

Verify the new kernel parameter values by running the following on both Oracle RAC nodes in the cluster:

# /sbin/sysctl -a | grep shm
vm.hugetlb_shm_group = 0
kernel.shmmni = 4096
kernel.shmall = 268435456
kernel.shmmax = 4294967295

# /sbin/sysctl -a | grep sem
kernel.sem = 250        32000   100     128

# /sbin/sysctl -a | grep file-max
fs.file-max = 65536

# /sbin/sysctl -a | grep ip_local_port_range
net.ipv4.ip_local_port_range = 1024     65000

# /sbin/sysctl -a | grep 'core\.[rw]mem'
net.core.rmem_default = 1048576
net.core.wmem_default = 262144
net.core.rmem_max = 1048576
net.core.wmem_max = 262144

Setting Shell Limits for the oracle User

To improve the performance of the software on Linux systems, Oracle recommends you increase the following shell limits for the oracle user:

To make these changes, run the following as root:

cat >> /etc/security/limits.conf <<EOF
oracle soft nproc 2047
oracle hard nproc 16384
oracle soft nofile 1024
oracle hard nofile 65536
EOF

cat >> /etc/pam.d/login <<EOF
session required /lib/security/pam_limits.so
EOF

Update the default shell startup file for the "oracle" UNIX account.

• For the Bourne, Bash, or Korn shell, add the following lines to the /etc/profile file by running the following command:

  cat >> /etc/profile <<EOF
  if [ \$USER = "oracle" ]; then 
   if [ \$SHELL = "/bin/ksh" ]; then
     ulimit -p 16384
     ulimit -n 65536
   else
     ulimit -u 16384 -n 65536
   fi
   umask 022
  fi
  EOF

• For the C shell (csh or tcsh), add the following lines to the /etc/csh.login file by running the following command:

  cat >> /etc/csh.login <<EOF
  if ( \$USER == "oracle" ) then
   limit maxproc 16384
   limit descriptors 65536
  endif
  EOF

Setting the Correct Date and Time on All Cluster Nodes

During the installation of Oracle Clusterware, the Database, and the Companion CD, the Oracle Universal Installer (OUI) first installs the software to the local node running the installer (i.e. linux1). The software is then copied remotely to all of the remaining nodes in the cluster (i.e. linux2). During the remote copy process, the OUI will execute the UNIX "tar" command on each of the remote nodes to extract the files that were archived and copied over. If the date and time on the node performing the install is greater than that of the node it is copying to, the OUI will throw an error from the "tar" command indicating it is attempting to extract files stamped with a time in the future:

Error while copying directory 
    /u01/app/crs with exclude file list 'null' to nodes 'linux2'.
[PRKC-1002 : All the submitted commands did not execute successfully]
---------------------------------------------
linux2:
   /bin/tar: ./bin/lsnodes: time stamp 2009-07-28 09:21:34 is 735 s in the future
   /bin/tar: ./bin/olsnodes: time stamp 2009-07-28 09:21:34 is 735 s in the future
   ...(more errors on this node)

Please note that although this would seem like a severe error from the OUI, it can safely be disregarded as a warning. The "tar" command DOES actually extract the files; however, when you perform a listing of the files (using ls -l) on the remote node, they will be missing the time field until the time on the server is greater than the timestamp of the file.

Before starting any of the above noted installations, ensure that each member node of the cluster is set as closely as possible to the same date and time. Oracle strongly recommends using the Network Time Protocol feature of most operating systems for this purpose, with both Oracle RAC nodes using the same reference Network Time Protocol server.

Accessing a Network Time Protocol server, however, may not always be an option. In this case, when manually setting the date and time for the nodes in the cluster, ensure that the date and time of the node you are performing the software installations from (linux1) is less than all other nodes in the cluster (linux2). I generally use a 20 second difference as shown in the following example:

Setting the date and time from linux1:

# date -s "7/28/2009 23:00:00"

Setting the date and time from linux2:

# date -s "7/28/2009 23:00:20"

The two-node RAC configuration described in this article does not make use of a Network Time Protocol server.

Perform the following configuration procedures on both Oracle RAC nodes in the cluster!

Oracle9i Release 1 (9.0.1) and Oracle9i Release 2 ( 9.2.0.1) used a userspace watchdog daemon called watchdogd to monitor the health of the cluster and to restart a RAC node in case of a failure. Starting with Oracle9i Release 2 (9.2.0.2) (and still available in Oracle 10g Release 2), the watchdog daemon has been deprecated by a Linux kernel module named hangcheck-timer which addresses availability and reliability problems much better. The hang-check timer is loaded into the Linux kernel and checks if the system hangs. It will set a timer and check the timer after a certain amount of time. There is a configurable threshold to hang-check that, if exceeded will reboot the machine. Although the hangcheck-timer module is not required for Oracle Clusterware (Cluster Manager) operation, it is highly recommended by Oracle.

The hangcheck-timer.ko Module

The hangcheck-timer module uses a kernel-based timer that periodically checks the system task scheduler to catch delays in order to determine the health of the system. If the system hangs or pauses, the timer resets the node. The hangcheck-timer module uses the Time Stamp Counter (TSC) CPU register, which is incremented at each clock signal. The TCS offers much more accurate time measurements because this register is updated by the hardware automatically.

Much more information about the hangcheck-timer project can be found here.

Installing the hangcheck-timer.ko Module

The hangcheck-timer was normally shipped only by Oracle, however, this module is now included with Red Hat Linux AS starting with kernel versions 2.4.9-e.12 and higher. The hangcheck-timer should already be included. Use the following to ensure that you have the module included:

# find /lib/modules -name "hangcheck-timer.ko"
/lib/modules/2.6.18-128.el5/kernel/drivers/char/hangcheck-timer.ko

Configuring and Loading the hangcheck-timer Module

There are two key parameters to the hangcheck-timer module:

• hangcheck-tick: This parameter defines the period of time between checks of system health. The default value is 60 seconds; Oracle recommends setting it to 30 seconds.
• hangcheck-margin: This parameter defines the maximum hang delay that should be tolerated before hangcheck-timer resets the RAC node. It defines the margin of error in seconds. The default value is 180 seconds; Oracle recommends setting it to 180 seconds.

system hang time > (hangcheck_tick + hangcheck_margin)

Each time the hangcheck-timer kernel module is loaded (manually or by Oracle), it needs to know what value to use for each of the two parameters we just discussed: (hangcheck-tick and hangcheck-margin). These values need to be available after each reboot of the Linux server. To do that, make an entry with the correct values to the /etc/modprobe.conf file as follows:

# su -
# echo "options hangcheck-timer hangcheck_tick=30 hangcheck_margin=180" >> /etc/modprobe.conf

Manually Loading the Hangcheck Kernel Module for Testing

Oracle is responsible for loading the hangcheck-timer kernel module when required. For that reason, it is not required to perform a modprobe or insmod of the hangcheck-timer kernel module in any of the startup files (i.e. /etc/rc.local).

It is only out of pure habit that I continue to include a modprobe of the hangcheck-timer kernel module in the /etc/rc.local file. Someday I will get over it, but realize that it does not hurt to include a modprobe of the hangcheck-timer kernel module during startup.

So to keep myself sane and able to sleep at night, I always configure the loading of the hangcheck-timer kernel module on each startup as follows:

# echo "/sbin/modprobe hangcheck-timer" >> /etc/rc.local

Note: You don't have to manually load the hangcheck-timer kernel module using modprobe or insmod after each reboot. The hangcheck-timer module will be loaded by Oracle automatically when needed.

Now, to test the hangcheck-timer kernel module to verify it is picking up the correct parameters we defined in the /etc/modprobe.conf file, use the modprobe command. Although you could load the hangcheck-timer kernel module by passing it the appropriate parameters (e.g. insmod hangcheck-timer hangcheck_tick=30 hangcheck_margin=180), we want to verify that it is picking up the options we set in the /etc/modprobe.conf file.

To manually load the hangcheck-timer kernel module and verify it is using the correct values defined in the /etc/modprobe.conf file, run the following command:

# su -
# modprobe hangcheck-timer
# grep Hangcheck /var/log/messages | tail -2
Jul 30 17:50:43 linux1 kernel: Hangcheck: starting hangcheck timer 0.9.0 (tick is 30 seconds, margin is 180 seconds).
Jul 30 17:50:43 linux1 kernel: Hangcheck: Using get_cycles().

Perform the following configuration procedures on both Oracle RAC nodes in the cluster!

Before you can install Oracle RAC 10g, you must configure secure shell (SSH) for the UNIX user account you plan to use to install Oracle Clusterware 10g and the Oracle Database 10g software. The installation and configuration tasks described in this section will need to be performed on both Oracle RAC nodes. As configured earlier in this article, the software owner for Oracle Clusterware 10g and the Oracle Database 10g software will be "oracle".

The goal here is to setup user equivalence for the oracle UNIX user account. User equivalence enables the oracle UNIX user account to access all other nodes in the cluster (running commands and copying files) without the need for a password. Oracle added support in 10g Release 1 for using the SSH tool suite for setting up user equivalence. Before Oracle Database 10g, user equivalence had to be configured using remote shell (RSH).

The SSH configuration described in this article uses SSH1. If SSH is not available, then OUI attempts to use rsh and rcp instead. These services, however, are disabled by default on most Linux systems. The use of RSH will not be discussed in this article.

You need either an RSA or a DSA key for the SSH protocol. RSA is used with the SSH 1.5 protocol, while DSA is the default for the SSH 2.0 protocol. With OpenSSH, you can use either RSA or DSA. For the purpose of this article, we will configure SSH using SSH1.

Note: If you have an SSH2 installation, and you cannot use SSH1, then refer to your SSH distribution documentation to configure SSH1 compatibility or to configure SSH2 with DSA. This type of configuration is beyond the scope of this article and will not be discussed.

So, why do we have to setup user equivalence? Installing Oracle Clusterware and the Oracle Database software is only performed from one node in a RAC cluster. When running the Oracle Universal Installer (OUI) on that particular node, it will use the ssh and scp commands to run remote commands on and copy files (the Oracle software) to all other nodes within the RAC cluster. The oracle UNIX user account on the node running the OUI (runInstaller) must be trusted by all other nodes in your RAC cluster. This means that you must be able to run the secure shell commands (ssh or scp) on the Linux server you will be running the OUI from against all other Linux servers in the cluster without being prompted for a password.

Please note that the use of secure shell is not required for normal RAC operation. This configuration, however, must to be enabled for RAC and patchset installations as well as creating the clustered database.

The methods required for configuring SSH1, an RSA key, and user equivalence is described in the following sections.

Configuring the Secure Shell

To determine if SSH is installed and running, enter the following command:

# pgrep sshd
2400

If SSH is running, then the response to this command is a list of process ID number(s). Run this command on both Oracle RAC nodes in the cluster to verify the SSH daemons are installed and running!

To find out more about SSH, refer to the man page:

# man ssh

Creating the RSA Keys on Both Oracle RAC Nodes

The first step in configuring SSH is to create an RSA public/private key pair on both Oracle RAC nodes in the cluster. The command to do this will create a public and private key for RSA (for a total of two keys per node). The content of the RSA public keys will then need to be copied into an authorized key file which is then distributed to both Oracle RAC nodes in the cluster.

Use the following steps to create the RSA key pair. Please note that these steps will need to be completed on both Oracle RAC nodes in the cluster:

1. Log on as the oracle UNIX user account.

   # su - oracle

2. If necessary, create the .ssh directory in the oracle user's home directory and set the correct permissions to ensure that only the oracle user has read and write permissions:

   $ mkdir -p ~/.ssh
   $ chmod 700 ~/.ssh

3. Enter the following command to generate an RSA key pair (public and private key) for the SSH protocol:

   $ /usr/bin/ssh-keygen -t rsa

   At the prompts:

   • Accept the default location for the key files (press [ENTER]).
   • Enter and confirm a pass phrase. This should be different from the oracle UNIX user account password however it is not a requirement.

   This command will write the public key to the ~/.ssh/id_rsa.pub file and the private key to the ~/.ssh/id_rsa file. Note that you should never distribute the private key to anyone!

4. Repeat the above steps for both Oracle RAC nodes in the cluster.

Now that both Oracle RAC nodes contain a public and private key for RSA, you will need to create an authorized key file on one of the nodes. An authorized key file is nothing more than a single file that contains a copy of everyone's (every node's) RSA public key. Once the authorized key file contains all of the public keys, it is then distributed to all other nodes in the cluster.

Complete the following steps on one of the nodes in the cluster to create and then distribute the authorized key file. For the purpose of this article, I am using linux1:

1. First, determine if an authorized key file already exists on the node (~/.ssh/authorized_keys). In most cases this will not exist since this article assumes you are working with a new install. If the file doesn't exist, create it now:

   $ touch ~/.ssh/authorized_keys
   $ cd ~/.ssh
   $ ls -l *.pub
   -rw-r--r-- 1 oracle oinstall 395 Jul 30 18:51 id_rsa.pub

   The listing above should show the id_rsa.pub public key created in the previous section.

2. In this step, use SCP (Secure Copy) or SFTP (Secure FTP) to copy the content of the ~/.ssh/id_rsa.pub public key from both Oracle RAC nodes in the cluster to the authorized key file just created (~/.ssh/authorized_keys). Again, this will be done from linux1. You will be prompted for the oracle UNIX user account password for both Oracle RAC nodes accessed.

   The following example is being run from linux1 and assumes a two-node cluster, with nodes linux1 and linux2:

   $ ssh linux1 cat ~/.ssh/id_rsa.pub >> ~/.ssh/authorized_keys
   The authenticity of host 'linux1 (192.168.1.100)' can't be established.
   RSA key fingerprint is 46:0f:1b:ac:a6:9d:86:4d:38:45:85:76:ad:3b:e7:c9.
   Are you sure you want to continue connecting (yes/no)? yes
   Warning: Permanently added 'linux1,192.168.1.100' (RSA) to the list of known hosts.
   oracle@linux1's password: xxxxx
   
   $ ssh linux2 cat ~/.ssh/id_rsa.pub >> ~/.ssh/authorized_keys
   The authenticity of host 'linux2 (192.168.1.101)' can't be established.
   RSA key fingerprint is eb:39:98:a7:d9:61:02:a2:80:3b:75:71:70:42:d0:07.
   Are you sure you want to continue connecting (yes/no)? yes
   Warning: Permanently added 'linux2,192.168.1.101' (RSA) to the list of known hosts.
   oracle@linux2's password: xxxxx

   Note: The first time you use SSH to connect to a node from a particular system, you will see a message similar to the following:

   The authenticity of host 'linux1 (192.168.1.100)' can't be established.
   RSA key fingerprint is 46:0f:1b:ac:a6:9d:86:4d:38:45:85:76:ad:3b:e7:c9.
   Are you sure you want to continue connecting (yes/no)? yes

   Enter yes at the prompt to continue. You will not see this message again when you connect from this system to the same node.

3. At this point, we have the RSA public key from every node in the cluster in the authorized key file (~/.ssh/authorized_keys) on linux1. We now need to copy it to the remaining nodes in the cluster. In our two-node cluster example, the only remaining node is linux2. Use the scp command to copy the authorized key file to all remaining nodes in the cluster:

   $ scp ~/.ssh/authorized_keys linux2:.ssh/authorized_keys
   oracle@linux2's password: xxxxx
   authorized_keys                 100%  790     0.8KB/s   00:00

4. Change the permission of the authorized key file for both Oracle RAC nodes in the cluster by logging into the node and running the following:

   $ chmod 600 ~/.ssh/authorized_keys

5. At this point, if you use ssh to log in to or run a command on another node, you are prompted for the pass phrase that you specified when you created the RSA key. For example, test the following from linux1:

   $ ssh linux1 hostname
   Enter passphrase for key '/home/oracle/.ssh/id_rsa': xxxxx
   linux1
   
   $ ssh linux2 hostname
   Enter passphrase for key '/home/oracle/.ssh/id_rsa': xxxxx
   linux2

   Note: If you see any other messages or text, apart from the host name, then the Oracle installation will fail. Make any changes required to ensure that only the host name is displayed when you enter these commands. You should ensure that any part of a login script that generates any output, or asks any questions, is modified so it acts only when the shell is an interactive shell.

Verify that the following startup commands are included on both of the Oracle RAC nodes in the cluster!

Up to this point, we have talked in great detail about the parameters and resources that need to be configured on both nodes in the Oracle RAC 10g configuration. This section will review those parameters, commands, and entries (in previous sections of this document) that need to occur on both Oracle RAC nodes when they are booted.

For each of the startup files below, entries in gray should be included in each startup file.

/etc/modprobe.conf

(All parameters and values to be used by kernel modules.)

.................................................................
alias eth0 r8169
alias eth1 e1000
alias scsi_hostadapter ata_piix
alias snd-card-0 snd-intel8x0
options snd-card-0 index=0
options snd-intel8x0 index=0
remove snd-intel8x0 { /usr/sbin/alsactl store 0 >/dev/null 2>&1 || : ; }; /sbin/modprobe -r --ignore-remove snd-intel8x0
options hangcheck-timer hangcheck_tick=30 hangcheck_margin=180
.................................................................

/etc/sysctl.conf

(We wanted to adjust the default and maximum send buffer size as well as the default and maximum receive buffer size for the interconnect. This file also contains those parameters responsible for configuring shared memory, semaphores, file handles, and local IP range for use by the Oracle instance.)

.................................................................
# Kernel sysctl configuration file for Oracle Enterprise Linux
#
# For binary values, 0 is disabled, 1 is enabled.  See sysctl(8) and
# sysctl.conf(5) for more details.

# Controls IP packet forwarding
net.ipv4.ip_forward = 0

# Controls source route verification
net.ipv4.conf.default.rp_filter = 1

# Do not accept source routing
net.ipv4.conf.default.accept_source_route = 0

# Controls the System Request debugging functionality of the kernel
kernel.sysrq = 0

# Controls whether core dumps will append the PID to the core filename
# Useful for debugging multi-threaded applications
kernel.core_uses_pid = 1

# Controls the use of TCP syncookies
net.ipv4.tcp_syncookies = 1

# Controls the maximum size of a message, in bytes
kernel.msgmnb = 65536

# Controls the default maxmimum size of a mesage queue
kernel.msgmax = 65536

# +---------------------------------------------------------+
# | ADJUSTING NETWORK SETTINGS                              |
# +---------------------------------------------------------+
# | With Oracle 9.2.0.1 and onwards, Oracle now makes use   |
# | of UDP as the default protocol on Linux for             |
# | inter-process communication (IPC), such as Cache Fusion |
# | and Cluster Manager buffer transfers between instances  |
# | within the RAC cluster. Oracle strongly suggests to     |
# | adjust the default and maximum receive buffer size      |
# | (SO_RCVBUF socket option) to 1024 MB, and the default   |
# | and maximum send buffer size (SO_SNDBUF socket option)  |
# | to 256 KB. The receive buffers are used by TCP and UDP  |
# | to hold received data until it is read by the           |
# | application. The receive buffer cannot overflow because |
# | the peer is not allowed to send data beyond the buffer  |
# | size window. This means that datagrams will be          |
# | discarded if they don't fit in the socket receive       |
# | buffer. This could cause the sender to overwhelm the    |
# | receiver.                                               |
# +---------------------------------------------------------+

# +---------------------------------------------------------+
# | Default setting in bytes of the socket "receive" buffer |
# | which may be set by using the SO_RCVBUF socket option.  |
# +---------------------------------------------------------+
net.core.rmem_default=1048576

# +---------------------------------------------------------+
# | Maximum setting in bytes of the socket "receive" buffer |
# | which may be set by using the SO_RCVBUF socket option.  |
# +---------------------------------------------------------+
net.core.rmem_max=1048576

# +---------------------------------------------------------+
# | Default setting in bytes of the socket "send" buffer    |
# | which may be set by using the SO_SNDBUF socket option.  |
# +---------------------------------------------------------+
net.core.wmem_default=262144

# +---------------------------------------------------------+
# | Maximum setting in bytes of the socket "send" buffer    |
# | which may be set by using the SO_SNDBUF socket option.  |
# +---------------------------------------------------------+
net.core.wmem_max=262144

# +---------------------------------------------------------+
# | ADJUSTING ADDITIONAL KERNEL PARAMETERS FOR ORACLE       |
# +---------------------------------------------------------+
# | Configure the kernel parameters for all Oracle Linux    |
# | servers by setting shared memory and semaphores,        |
# | setting the maximum amount of file handles, and setting |
# | the IP local port range.                                |
# +---------------------------------------------------------+

# +---------------------------------------------------------+
# | SHARED MEMORY                                           |
# +---------------------------------------------------------+
# Controls the maximum shared segment size, in bytes
kernel.shmmax = 4294967295

# Controls the maximum number of shared memory segments, in pages
kernel.shmall = 268435456

# Controls the maximum number of shared memory segments system wide
kernel.shmmni = 4096

# +---------------------------------------------------------+
# | SEMAPHORES                                              |
# | ----------                                              |
# |                                                         |
# | SEMMSL_value  SEMMNS_value  SEMOPM_value  SEMMNI_value  |
# |                                                         |
# +---------------------------------------------------------+
kernel.sem=250 32000 100 128

# +---------------------------------------------------------+
# | FILE HANDLES                                            |
# ----------------------------------------------------------+
fs.file-max=65536

# +---------------------------------------------------------+
# | LOCAL IP RANGE                                          |
# ----------------------------------------------------------+
net.ipv4.ip_local_port_range=1024 65000
.................................................................

Note: Verify that each of the required kernel parameters (above) are configured in the /etc/sysctl.conf file. Then, ensure that each of these parameters are truly in effect by running the following command on both Oracle RAC nodes in the cluster:

# sysctl -p
net.ipv4.ip_forward = 0
net.ipv4.conf.default.rp_filter = 1
net.ipv4.conf.default.accept_source_route = 0
kernel.sysrq = 0
kernel.core_uses_pid = 1
net.ipv4.tcp_syncookies = 1
kernel.msgmnb = 65536
kernel.msgmax = 65536
net.core.rmem_default = 1048576
net.core.rmem_max = 1048576
net.core.wmem_default = 262144
net.core.wmem_max = 262144
kernel.shmmax = 4294967295
kernel.shmall = 268435456
kernel.shmmni = 4096
kernel.sem = 250 32000 100 128
fs.file-max = 65536
net.ipv4.ip_local_port_range = 1024 65000

/etc/hosts

(All machine/IP entries for nodes in our RAC cluster.)

.................................................................
# Do not remove the following line, or various programs
# that require network functionality will fail.

127.0.0.1        localhost.localdomain   localhost

# Public Network - (eth0)
192.168.1.100    linux1
192.168.1.101    linux2

# Private Interconnect - (eth1)
192.168.2.100    linux1-priv
192.168.2.101    linux2-priv

# Public Virtual IP (VIP) addresses - (eth0:1)
192.168.1.200    linux1-vip
192.168.1.201    linux2-vip

# Private Storage Network for Openfiler - (eth1)
192.168.1.195    openfiler1
192.168.2.195    openfiler1-priv

# Miscellaneous Nodes
192.168.1.1      router
192.168.1.102    alex
192.168.1.103    nascar
192.168.1.105    packmule
192.168.1.106    melody
192.168.1.120    cartman
192.168.1.121    domo
192.168.1.122    switch1
192.168.1.190    george
192.168.1.245    accesspoint
.................................................................

/etc/udev/rules.d/55-openiscsi.rules

.................................................................
# /etc/udev/rules.d/55-openiscsi.rules
KERNEL=="sd*", BUS=="scsi", PROGRAM="/etc/udev/scripts/iscsidev.sh %b",SYMLINK+="iscsi/%c/part%n"
.................................................................

/etc/udev/scripts/iscsidev.sh

.................................................................
#!/bin/sh

# FILE: /etc/udev/scripts/iscsidev.sh

BUS=${1}
HOST=${BUS%%:*}

[ -e /sys/class/iscsi_host ] || exit 1

file="/sys/class/iscsi_host/host${HOST}/device/session*/iscsi_session*/targetname"

target_name=$(cat ${file})

# This is not an open-scsi drive
if [ -z "${target_name}" ]; then
   exit 1
fi

echo "${target_name##*.}"
.................................................................

/etc/rc.local

(Loading the hangcheck-timer kernel module.)

.................................................................
#!/bin/sh
#
# This script will be executed *after* all the other init scripts.
# You can put your own initialization stuff in here if you don't
# want to do the full Sys V style init stuff.

touch /var/lock/subsys/local

# +---------------------------------------------------------+
# | HANGCHECK TIMER                                         |
# | (I do not believe this is required, but doesn't hurt)   |
# +---------------------------------------------------------+

/sbin/modprobe hangcheck-timer
.................................................................

Most of the installation and configuration procedures in this section should be performed on both Oracle RAC nodes in the cluster! Creating the OCFS2 filesystem, however, should only be executed on one of nodes in the RAC cluster.

It is now time to install and configure the Oracle Cluster File System, Release 2 (OCFS2) software. Developed by Oracle Corporation, OCFS2 is a Cluster File System which allows all nodes in a cluster to concurrently access a device via the standard file system interface. This allows for easy management of applications that need to run across a cluster.

OCFS Release 1 was released in December 2002 to enable Oracle Real Application Cluster (RAC) users to run the clustered database without having to deal with RAW devices. The file system was designed to store database related files, such as data files, control files, redo logs, archive logs, etc. OCFS2 is the next generation of the Oracle Cluster File System. It has been designed to be a general purpose cluster file system. With it, users can store not only database related files on a shared disk, but also store Oracle binaries and configuration files (a shared Oracle Home for example) making management of RAC even easier.

In this guide, you will be using the release of OCFS2 included with Oracle Enterprise Linux Release 5.3 (OCFS2 Release 1.2.9-1) to store the two files that are required to be shared by the Oracle Clusterware software. Along with these two files, you will also be using this space to store the shared SPFILE for all Oracle ASM instances.

See this page for more information on OCFS2 (including Installation Notes) for Linux.

Install OCFS2

In previous editions of this article, this would be the time where you would need to download the OCFS2 software from http://oss.oracle.com/. This is no longer necessary since the OCFS2 software is included with Oracle Enterprise Linux. The OCFS2 software stack includes the following packages:

32-bit (x86) Installations

• OCFS2 Kernel Driver
• ocfs2-x.x.x-x.el5-x.x.x-x.el5.i686.rpm - (for default kernel)
• ocfs2-x.x.x-x.el5PAE-x.x.x-x.el5.i686.rpm - (for PAE kernel)
• ocfs2-x.x.x-x.el5xen-x.x.x-x.el5.i686.rpm - (for xen kernel)
• OCFS2 Tools
• ocfs2-tools-x.x.x-x.el5.i386.rpm
• OCFS2 Tools Development
• ocfs2-tools-devel-x.x.x-x.el5.i386.rpm
• OCFS2 Console
• ocfs2console-x.x.x-x.el5.i386.rpm

64-bit (x86_64) Installations

• OCFS2 Kernel Driver
• ocfs2-x.x.x-x.el5-x.x.x-x.el5.x86_64.rpm - (for default kernel)
• ocfs2-x.x.x-x.el5PAE-x.x.x-x.el5.x86_64.rpm - (for PAE kernel)
• ocfs2-x.x.x-x.el5xen-x.x.x-x.el5.x86_64.rpm - (for xen kernel)
• OCFS2 Tools
• ocfs2-tools-x.x.x-x.el5.x86_64.rpm
• OCFS2 Tools Development
• ocfs2-tools-devel-x.x.x-x.el5.x86_64.rpm
• OCFS2 Console
• ocfs2console-x.x.x-x.el5.x86_64.rpm

With Oracle Enterprise Linux 5.3, the OCFS2 software packages do not get installed by default. The OCFS2 software packages can be found on CD #3. To determine if the OCFS2 packages are installed (which in most cases, they will not be), perform the following on both Oracle RAC nodes:

# rpm -qa | grep ocfs2 | sort

If the OCFS2 packages are not installed, load the Oracle Enterprise Linux CD #3 into each of the Oracle RAC nodes and perform the following:

From Oracle Enterprise Linux 5 - [CD #3]
# mount -r /dev/cdrom /media/cdrom
# cd /media/cdrom/Server
# rpm -Uvh ocfs2-tools-1.2.7-1.el5.i386.rpm
# rpm -Uvh ocfs2-2.6.18-128.el5-1.2.9-1.el5.i686.rpm
# rpm -Uvh ocfs2console-1.2.7-1.el5.i386.rpm
# cd /
# eject

After installing the OCFS2 packages, verify from both Oracle RAC nodes that the software is installed:

# rpm -qa | grep ocfs2 | sort
ocfs2-2.6.18-128.el5-1.2.9-1.el5
ocfs2console-1.2.7-1.el5
ocfs2-tools-1.2.7-1.el5

Disable SELinux (RHEL4 U2 and higher)

Users of RHEL4 U2 and higher (Oracle Enterprise Linux 5.3 is based on RHEL 5.3) are advised that OCFS2 currently does not work with SELinux enabled. If you are using RHEL4 U2 or higher (which includes us since we are using Oracle Enterprise Linux 5.3) you will need to verify SELinux is disabled in order for the O2CB service to execute.

During the installation of Oracle Enterprise Linux, we Disabled SELinux on the SELinux screen. If, however, you did not disable SELinux during the installation phase, you can use the tool system-config-securitylevel to disable SELinux.

To disable SELinux (or verify SELinux is disabled), run the "Security Level Configuration" GUI utility:

# /usr/bin/system-config-securitylevel &

This will bring up the following screen:

Figure 16: Security Level Configuration Opening Screen / Firewall Disabled

Now, click the SELinux tab and select the "Disabled" option. After clicking [OK], you will be presented with a warning dialog. Simply acknowledge this warning by clicking "Yes". Your screen should now look like the following after disabling the SELinux option:

Figure 17: SELinux Disabled

If you needed to disable SELinux in this section on any of the nodes, those nodes will need to be rebooted to implement the change. SELinux must be disabled before you can continue with configuring OCFS2!

Configure OCFS2

OCFS2 will be configured to use the private network (192.168.2.0) for all of its network traffic as recommended by Oracle. While OCFS2 does not take much bandwidth, it does require the nodes to be alive on the network and sends regular keepalive packets to ensure that they are. To avoid a network delay being interpreted as a node disappearing on the net which could lead to a node-self-fencing, a private interconnect is recommended. It is safe to use the same private interconnect for both Oracle RAC and OCFS2.

A popular question then is what node name should be used and should it be related to the IP address? The node name needs to match the hostname of the machine. The IP address need not be the one associated with that hostname. In other words, any valid IP address on that node can be used. OCFS2 will not attempt to match the node name (hostname) with the specified IP address.

The next step is to generate and configure the /etc/ocfs2/cluster.conf file on both Oracle RAC nodes in the cluster. The easiest way to accomplish this is to run the GUI tool ocfs2console. In this section, we will not only create and configure the /etc/ocfs2/cluster.conf file using ocfs2console, but will also create and start the cluster stack O2CB. When the /etc/ocfs2/cluster.conf file is not present, (as will be the case in our example), the ocfs2console tool will create this file along with a new cluster stack service (O2CB) with a default cluster name of ocfs2. This will need to be done on both Oracle RAC nodes in the cluster as the root user account:

$ su -
# ocfs2console &

This will bring up the GUI as shown below:

Figure 18: ocfs2console GUI

Using the ocfs2console GUI tool, perform the following steps:

1. Select [Cluster] -> [Configure Nodes...]. This will start the OCFS2 Cluster Stack (Figure 19) and bring up the "Node Configuration" dialog.
2. On the "Node Configuration" dialog, click the [Add] button.
   • This will bring up the "Add Node" dialog.
   • In the "Add Node" dialog, enter the Host name and IP address for the first node in the cluster. Leave the IP Port set to its default value of 7777. In my example, I added both nodes using linux1 / 192.168.2.100 for the first node and linux2 / 192.168.2.101 for the second node.
     Note: The node name you enter "must" match the hostname of the machine and the IP addresses will use the private interconnect.
   • Click [Apply] on the "Node Configuration" dialog - All nodes should now be "Active" as shown in Figure 20.
   • Click [Close] on the "Node Configuration" dialog.
3. After verifying all values are correct, exit the application using [File] -> [Quit]. This needs to be performed on both Oracle RAC nodes in the cluster.

Figure 19: Starting the OCFS2 Cluster Stack

The following dialog show the OCFS2 settings I used for the nodes linux1 and linux2:

Figure 20: Configuring Nodes for OCFS2

Note: See the Troubleshooting section if you get the error:

    o2cb_ctl: Unable to access cluster service while creating node

After exiting the ocfs2console, you will have a /etc/ocfs2/cluster.conf similar to the following. This process needs to be completed on both Oracle RAC nodes in the cluster and the OCFS2 configuration file should be exactly the same for all of the nodes:

node:
        ip_port = 7777
        ip_address = 192.168.2.100
        number = 0
        name = linux1
        cluster = ocfs2

node:
        ip_port = 7777
        ip_address = 192.168.2.101
        number = 1
        name = linux2
        cluster = ocfs2

cluster:
        node_count = 2
        name = ocfs2

O2CB Cluster Service

Before we can do anything with OCFS2 like formatting or mounting the file system, we need to first have OCFS2's cluster stack, O2CB, running (which it will be as a result of the configuration process performed above). The stack includes the following services:

• NM: Node Manager that keep track of all the nodes in the cluster.conf
• HB: Heart beat service that issues up/down notifications when nodes join or leave the cluster
• TCP: Handles communication between the nodes
• DLM: Distributed lock manager that keeps track of all locks, its owners and status
• CONFIGFS: User space driven configuration file system mounted at /config
• DLMFS: User space interface to the kernel space DLM

All of the above cluster services have been packaged in the o2cb system service (/etc/init.d/o2cb). Here is a short listing of some of the more useful commands and options for the o2cb system service.

Note: The following commands are for documentation purposes only and do not need to be run when installing and configuring OCFS2 for this article!

• /etc/init.d/o2cb status

  Module "configfs": Loaded
  Filesystem "configfs": Mounted
  Module "ocfs2_nodemanager": Loaded
  Module "ocfs2_dlm": Loaded
  Module "ocfs2_dlmfs": Loaded
  Filesystem "ocfs2_dlmfs": Mounted
  Checking O2CB cluster ocfs2: Online
    Heartbeat dead threshold: 31
    Network idle timeout: 30000
    Network keepalive delay: 2000
    Network reconnect delay: 2000
  Checking O2CB heartbeat: Not active

• /etc/init.d/o2cb offline ocfs2

  Stopping O2CB cluster ocfs2: OK

  The above command will offline the cluster we created, ocfs2.

• /etc/init.d/o2cb unload

  Unmounting ocfs2_dlmfs filesystem: OK
  Unloading module "ocfs2_dlmfs": OK
  Unmounting configfs filesystem: OK
  Unloading module "configfs": OK

  The above command will unload all OCFS2 modules.

• /etc/init.d/o2cb load

  Loading module "configfs": OK
  Mounting configfs filesystem at /sys/kernel/config: OK
  Loading module "ocfs2_nodemanager": OK
  Loading module "ocfs2_dlm": OK
  Loading module "ocfs2_dlmfs": OK
  Mounting ocfs2_dlmfs filesystem at /dlm: OK

  Loads all OCFS2 modules.

• /etc/init.d/o2cb online ocfs2

  Starting O2CB cluster ocfs2: OK

  The above command will online the cluster we created, ocfs2.

Configure O2CB to Start on Boot and Adjust O2CB Heartbeat Threshold

You now need to configure the on-boot properties of the OC2B driver so that the cluster stack services will start on each boot. You will also be adjusting the OCFS2 Heartbeat Threshold from its default setting of 31 to 61. Perform the following on both Oracle RAC nodes in the cluster:

# /etc/init.d/o2cb offline ocfs2
# /etc/init.d/o2cb unload
# /etc/init.d/o2cb configure
Configuring the O2CB driver.

This will configure the on-boot properties of the O2CB driver.
The following questions will determine whether the driver is loaded on
boot.  The current values will be shown in brackets ('[]').  Hitting
<ENTER> without typing an answer will keep that current value.  Ctrl-C
will abort.

Load O2CB driver on boot (y/n) [n]: y
Cluster to start on boot (Enter "none" to clear) [ocfs2]: ocfs2
Specify heartbeat dead threshold (>=7) [31]: 61
Specify network idle timeout in ms (>=5000) [30000]: 30000
Specify network keepalive delay in ms (>=1000) [2000]: 2000
Specify network reconnect delay in ms (>=2000) [2000]: 2000
Writing O2CB configuration: OK
Loading module "configfs": OK
Mounting configfs filesystem at /sys/kernel/config: OK
Loading module "ocfs2_nodemanager": OK
Loading module "ocfs2_dlm": OK
Loading module "ocfs2_dlmfs": OK
Mounting ocfs2_dlmfs filesystem at /dlm: OK
Starting O2CB cluster ocfs2: OK

Format the OCFS2 Filesystem

Note: Unlike the other tasks in this section, creating the OCFS2 file system should only be executed on one of nodes in the RAC cluster. I will be executing all commands in this section from linux1 only.

We can now start to make use of the iSCSI volume we partitioned for OCFS2 in the section "Create Partitions on iSCSI Volumes".

If the O2CB cluster is offline, start it. The format operation needs the cluster to be online, as it needs to ensure that the volume is not mounted on some other node in the cluster.

Earlier in this document, we created the directory /u02 under the section Create Mount Point for OCFS2 / Clusterware which will be used as the mount point for the OCFS2 cluster file system. This section contains the commands to create and mount the file system to be used for the Cluster Manager.

Note that it is possible to create and mount the OCFS2 file system using either the GUI tool ocfs2console or the command-line tool mkfs.ocfs2. From the ocfs2console utility, use the menu [Tasks] - [Format].

The instructions below demonstrate how to create the OCFS2 file system using the command-line tool mkfs.ocfs2.

To create the file system, we can use the Oracle executable mkfs.ocfs2. For the purpose of this example, I run the following command only from linux1 as the root user account using the local SCSI device name mapped to the iSCSI volume for crs — /dev/iscsi/crs/part1. Also note that I specified a label named "oracrsfiles" which will be referred to when mounting or un-mounting the volume:

$ su -
# mkfs.ocfs2 -b 4K -C 32K -N 4 -L oracrsfiles /dev/iscsi/crs/part1

mkfs.ocfs2 1.2.7
Filesystem label=oracrsfiles
Block size=4096 (bits=12)
Cluster size=32768 (bits=15)
Volume size=2145943552 (65489 clusters) (523912 blocks)
3 cluster groups (tail covers 977 clusters, rest cover 32256 clusters)
Journal size=67108864
Initial number of node slots: 4
Creating bitmaps: done
Initializing superblock: done
Writing system files: done
Writing superblock: done
Writing backup superblock: 1 block(s)
Formatting Journals: done
Writing lost+found: done
mkfs.ocfs2 successful

Mount the OCFS2 Filesystem

Now that the file system is created, we can mount it. Let's first do it using the command-line, then I'll show how to include it in the /etc/fstab to have it mount on each boot.

Note: Mounting the cluster file system will need to be performed on both Oracle RAC nodes in the cluster as the root user account using the OCFS2 label oracrsfiles!

First, here is how to manually mount the OCFS2 file system from the command-line. Remember that this needs to be performed as the root user account:

$ su -
# mount -t ocfs2 -o datavolume,nointr -L "oracrsfiles" /u02

If the mount was successful, you will simply get your prompt back. We should, however, run the following checks to ensure the file system is mounted correctly.

Use the mount command to ensure that the new file system is really mounted. This should be performed on both nodes in the RAC cluster:

# mount
/dev/mapper/VolGroup00-LogVol00 on / type ext3 (rw)
proc on /proc type proc (rw)
sysfs on /sys type sysfs (rw)
devpts on /dev/pts type devpts (rw,gid=5,mode=620)
/dev/hda1 on /boot type ext3 (rw)
tmpfs on /dev/shm type tmpfs (rw)
none on /proc/sys/fs/binfmt_misc type binfmt_misc (rw)
sunrpc on /var/lib/nfs/rpc_pipefs type rpc_pipefs (rw)
domo:Public on /domo type nfs (rw,addr=192.168.1.121)
configfs on /sys/kernel/config type configfs (rw)
ocfs2_dlmfs on /dlm type ocfs2_dlmfs (rw)
/dev/sde1 on /u02 type ocfs2 (rw,_netdev,datavolume,nointr,heartbeat=local)

Any other type of volume, including an Oracle home (which I will not be using for this article), should not be mounted with this mount option.

Why does it take so much time to mount the volume? It takes around 5 seconds for a volume to mount. It does so as to let the heartbeat thread stabilize. In a later release, Oracle plans to add support for a global heartbeat, which will make most mounts instant.

Configure OCFS2 to Mount Automatically at Startup

Let's take a look at what you've have done so far. You installed the OCFS2 software packages which will be used to store the shared files needed by Cluster Manager. After going through the install, you loaded the OCFS2 module into the kernel and then formatted the clustered file system. Finally, you mounted the newly created file system using the OCFS2 label "oracrsfiles". This section walks through the steps responsible for mounting the new OCFS2 file system each time the machine(s) are booted using its label.

Start by adding the following line to the /etc/fstab file on both Oracle RAC nodes in the cluster:

LABEL=oracrsfiles     /u02           ocfs2   _netdev,datavolume,nointr     0 0

Notice the "_netdev" option for mounting this file system. The _netdev mount option is a must for OCFS2 volumes. This mount option indicates that the volume is to be mounted after the network is started and dismounted before the network is shutdown.

Now, let's make sure that the ocfs2.ko kernel module is being loaded and that the file system will be mounted during the boot process.

If you have been following along with the examples in this article, the actions to load the kernel module and mount the OCFS2 file system should already be enabled. However, you should still check those options by running the following on both Oracle RAC nodes in the cluster as the root user account:

$ su -
# chkconfig --list o2cb
o2cb 0:off 1:off 2:on 3:on 4:on 5:on 6:off

Check Permissions on New OCFS2 Filesystem

Use the ls command to check ownership. The permissions should be set to 0775 with owner "oracle" and group "oinstall".

The following tasks only need to be executed on one of nodes in the RAC cluster. I will be executing all commands in this section from linux1 only.

Let's first check the permissions:

# ls -ld /u02
drwxr-xr-x 3 root root 4096 Jul 31 17:21 /u02

# chown oracle:oinstall /u02
# chmod 775 /u02

# ls -ld /u02
drwxrwxr-x 3 oracle oinstall 4096 Jul 31 17:21 /u02

Create Directory for Oracle Clusterware Files

The last mandatory task is to create the appropriate directory on the new OCFS2 file system that will be used for the Oracle Clusterware shared files. We will also modify the permissions of this new directory to allow the "oracle" owner and group "oinstall" read/write access.

The following tasks only need to be executed on one of nodes in the RAC cluster. I will be executing all commands in this section from linux1 only.

# mkdir -p /u02/oradata/racdb
# chown -R oracle:oinstall /u02/oradata
# chmod -R 775 /u02/oradata
# ls -l /u02/oradata
total 4
drwxrwxr-x 2 oracle oinstall 4096 Jul 31 17:31 racdb

Reboot Both Nodes

Before starting the next section, this would be a good place to reboot both of the nodes in the RAC cluster. When the machines come up, ensure that the cluster stack services are being loaded and the new OCFS2 file system is being mounted:

# mount
/dev/mapper/VolGroup00-LogVol00 on / type ext3 (rw)
proc on /proc type proc (rw)
sysfs on /sys type sysfs (rw)
devpts on /dev/pts type devpts (rw,gid=5,mode=620)
/dev/hda1 on /boot type ext3 (rw)
tmpfs on /dev/shm type tmpfs (rw)
none on /proc/sys/fs/binfmt_misc type binfmt_misc (rw)
sunrpc on /var/lib/nfs/rpc_pipefs type rpc_pipefs (rw)
configfs on /sys/kernel/config type configfs (rw)
ocfs2_dlmfs on /dlm type ocfs2_dlmfs (rw)
cartman:SHARE2 on /cartman type nfs (rw,addr=192.168.1.120)
/dev/sdc1 on /u02 type ocfs2 (rw,_netdev,datavolume,nointr,heartbeat=local)

If you modified the O2CB heartbeat threshold, you should verify that it is set correctly::

# cat /proc/fs/ocfs2_nodemanager/hb_dead_threshold
61

How to Determine OCFS2 Version

To determine which version of OCFS2 is running, use:

# cat /proc/fs/ocfs2/version
OCFS2 1.2.9 Wed Jan 21 21:32:59 EST 2009 (build 5e8325ec7f66b5189c65c7a8710fe8cb)

Most of the installation and configuration procedures in this section should be performed on both of the Oracle RAC nodes in the cluster! Creating the ASM disks, however, will only need to be performed on a single node within the cluster.

In this section, we will install and configure ASMLib 2.0 which will be used by Automatic Storage Management (ASM). In this article, we will use ASM as the shared file system and volume manager for all Oracle physical database files (data, online redo logs, control files, archived redo logs) as well as a Flash Recovery Area.

ASM was introduced in Oracle Database 10g Release 1 and is used to alleviate the DBA from having to manage individual files and drives. ASM is built into the Oracle kernel and provides the DBA with a way to manage thousands of disk drives 24x7 for both single and clustered instances of Oracle. All of the files and directories to be used for Oracle will be contained in a disk group. ASM automatically performs load balancing in parallel across all available disk drives to prevent hot spots and maximize performance, even with rapidly changing data usage patterns.

There are two different methods to configure ASM on Linux:

• ASM with ASMLib I/O: This method creates all Oracle database files on raw block devices managed by ASM using ASMLib calls. RAW devices are not required with this method as ASMLib works with block devices.
• ASM with Standard Linux I/O: This method creates all Oracle database files on raw character devices managed by ASM using standard Linux I/O system calls. You will be required to create RAW devices for all disk partitions used by ASM.

In this article, I will be using the "ASM with ASMLib I/O" method. Oracle states in Metalink Note 275315.1 that "ASMLib was provided to enable ASM I/O to Linux disks without the limitations of the standard UNIX I/O API". I plan on performing several tests in the future to identify the performance gains in using ASMLib. Those performance metrics and testing details are out of scope of this article and therefore will not be discussed.

If you would like to learn more about Oracle ASMLib 2.0, visit http://www.oracle.com/technology/tech/linux/asmlib/

Install ASMLib 2.0 Packages

In previous editions of this article, this would be the time where you would need to download the ASMLib 2.0 software from Oracle ASMLib Downloads for Red Hat Enterprise Linux Server 5. This is no longer necessary since the ASMLib software is included with Oracle Enterprise Linux (with the exception of the Userspace Library which is a separate download). The ASMLib 2.0 software stack includes the following packages:

32-bit (x86) Installations

• ASMLib Kernel Driver
• oracleasm-x.x.x-x.el5-x.x.x-x.el5.i686.rpm - (for default kernel)
• oracleasm-x.x.x-x.el5PAE-x.x.x-x.el5.i686.rpm - (for PAE kernel)
• oracleasm-x.x.x-x.el5xen-x.x.x-x.el5.i686.rpm - (for xen kernel)
• Userspace Library
• oracleasmlib-x.x.x-x.el5.i386.rpm
• Driver Support Files
• oracleasm-support-x.x.x-x.el5.i386.rpm

64-bit (x86_64) Installations

• ASMLib Kernel Driver
• oracleasm-x.x.x-x.el5-x.x.x-x.el5.x86_64.rpm - (for default kernel)
• oracleasm-x.x.x-x.el5PAE-x.x.x-x.el5.x86_64.rpm - (for PAE kernel)
• oracleasm-x.x.x-x.el5xen-x.x.x-x.el5.x86_64.rpm - (for xen kernel)
• Userspace Library
• oracleasmlib-x.x.x-x.el5.x86_64.rpm
• Driver Support Files
• oracleasm-support-x.x.x-x.el5.x86_64.rpm

With Oracle Enterprise Linux 5.3, the ASMLib 2.0 software packages do not get installed by default. The ASMLib 2.0 Kernel Drivers and Driver Support File can be found on CD #3. The Userspace Library will need to be downloaded as it is not included with Oracle Enterprise Linux. To determine if the Oracle ASMLib packages are installed (which in most cases, they will not be), perform the following on both Oracle RAC nodes:

# rpm -qa | grep oracleasm | sort

If the Oracle ASMLib 2.0 packages are not installed, load the Oracle Enterprise Linux CD #3 into each of the Oracle RAC nodes and perform the following:

From Oracle Enterprise Linux 5 - [CD #3]
# mount -r /dev/cdrom /media/cdrom
# cd /media/cdrom/Server
# rpm -Uvh oracleasm-support-2.1.2-1.el5.i386.rpm
# rpm -Uvh oracleasm-2.6.18-128.el5-2.0.5-1.el5.i686.rpm
# cd /
# eject

After installing the ASMLib packages, verify from both Oracle RAC nodes that the software is installed:

# rpm -qa | grep oracleasm | sort
oracleasm-2.6.18-128.el5-2.0.5-1.el5
oracleasm-support-2.1.2-1.el5

Getting Oracle ASMLib

As mentioned in the previous section, the ASMLib 2.0 software is included with Oracle Enterprise Linux with the exception of the Userspace Library (a.k.a. the ASMLib support library). The Userspace Library is required and can be downloaded for free at:

• oracleasmlib-2.0.4-1.el5.i386.rpm

After downloading the Userspace Library to both Oracle RAC nodes in the cluster, install it using the following:

# rpm -Uvh oracleasmlib-2.0.4-1.el5.i386.rpm
Preparing...                ########################################### [100%]
   1:oracleasmlib           ########################################### [100%]

For information on obtaining the ASMLib support library through the Unbreakable Linux Network (which is not a requirement for this article), please visit Getting Oracle ASMLib via the Unbreakable Linux Network.

Configuring and Loading the ASMLib 2.0 Packages

Now that you have installed the ASMLib Packages for Linux, you need to configure and load the ASM kernel module. This task needs to be run on both Oracle RAC nodes as the root user account:

$ su -
# /etc/init.d/oracleasm configure
Configuring the Oracle ASM library driver.

This will configure the on-boot properties of the Oracle ASM library
driver.  The following questions will determine whether the driver is
loaded on boot and what permissions it will have.  The current values
will be shown in brackets ('[]').  Hitting <ENTER> without typing an
answer will keep that current value.  Ctrl-C will abort.

Default user to own the driver interface []: oracle
Default group to own the driver interface []: oinstall
Start Oracle ASM library driver on boot (y/n) [n]: y
Scan for Oracle ASM disks on boot (y/n) [y]: y
Writing Oracle ASM library driver configuration: done
Initializing the Oracle ASMLib driver: [  OK  ]
Scanning the system for Oracle ASMLib disks: [  OK  ]

Create ASM Disks for Oracle

Creating the ASM disks only needs to be performed from one node in the RAC cluster as the root user account. I will be running these commands on linux1. On the other Oracle RAC node, you will need to perform a scandisk to recognize the new volumes. When that is complete, you should then run the oracleasm listdisks command on both Oracle RAC nodes to verify that all ASM disks were created and available.

In the section "Create Partitions on iSCSI Volumes", we configured (partitioned) four iSCSI volumes to be used by ASM. ASM will be used for storing Oracle database files like online redo logs, database files, control files, archived redo log files, and the flash recovery area. Use the local device names that were created by udev when configuring the four ASM volumns.

Note: If you are repeating this article using the same hardware (actually, the same shared logical drives), you may get a failure when attempting to create the ASM disks. If you do receive a failure, try listing all ASM disks that were used by the previous install using:

# /etc/init.d/oracleasm listdisks
VOL1
VOL2
VOL3
VOL4

As you can see, the results show that I have four ASM volumes already defined. If you have the four volumes already defined from a previous run, go ahead and remove them using the following commands. After removing the previously created volumes, use the "oracleasm createdisk" commands (below) to create the new volumes.

# /etc/init.d/oracleasm deletedisk VOL1
Removing ASM disk "VOL1" [  OK  ]
# /etc/init.d/oracleasm deletedisk VOL2
Removing ASM disk "VOL2" [  OK  ]
# /etc/init.d/oracleasm deletedisk VOL3
Removing ASM disk "VOL3" [  OK  ]
# /etc/init.d/oracleasm deletedisk VOL4
Removing ASM disk "VOL4" [  OK  ]

To create the ASM disks using the iSCSI target names to local device name mappings, type the following:

$ su -
# /etc/init.d/oracleasm createdisk VOL1 /dev/iscsi/asm1/part1
Marking disk "/dev/iscsi/asm1/part1" as an ASM disk [  OK  ]

# /etc/init.d/oracleasm createdisk VOL2 /dev/iscsi/asm2/part1
Marking disk "/dev/iscsi/asm2/part1" as an ASM disk [  OK  ] 

# /etc/init.d/oracleasm createdisk VOL3 /dev/iscsi/asm3/part1
Marking disk "/dev/iscsi/asm3/part1" as an ASM disk [  OK  ]

# /etc/init.d/oracleasm createdisk VOL4 /dev/iscsi/asm4/part1
Marking disk "/dev/iscsi/asm4/part1" as an ASM disk [  OK  ]

On all other nodes in the RAC cluster, you must perform a scandisk to recognize the new volumes:

# /etc/init.d/oracleasm scandisks
Scanning system for ASM disks [  OK  ]

We can now test that the ASM disks were successfully created by using the following command on both nodes in the RAC cluster as the root user account:

# /etc/init.d/oracleasm listdisks
VOL1
VOL2
VOL3
VOL4

The following download procedures only need to be performed on one node in the cluster!

The next step is to download and extract the required Oracle software packages from the Oracle Technology Network (OTN):

Note: If you do not currently have an account with OTN, you will need to create one. This is a FREE account!

Oracle offers a development and testing license free of charge. No support, however, is provided and the license does not permit production use. A full description of the license agreement is available on OTN.

32-bit (x86) Installations

http://www.oracle.com/technology/software/products/database/oracle10g/htdocs/10201linuxsoft.html

64-bit (x86_64) Installations

http://www.oracle.com/technology/software/products/database/oracle10g/htdocs/10201linx8664soft.html

You will be downloading and extracting the required software from Oracle to only one of the Linux nodes in the cluster—namely, linux1. You will perform all Oracle software installs from this machine. The Oracle installer will copy the required software packages to all other nodes in the RAC configuration using the remote access method we setup in Section 15 (Configure RAC Nodes for Remote Access using SSH).

Login to the node that you will be performing all of the Oracle installations from (linux1) as the "oracle" user account. In this example, you will be downloading the required Oracle software to linux1 and saving them to /home/oracle/orainstall.

Downloading and Extracting the Software

First, download the Oracle Clusterware Release 2 (10.2.0.1.0), Oracle Database 10g Release 2 (10.2.0.1.0), and Oracle Database 10g Companion CD Release 2 (10.2.0.1.0) software for either Linux x86 or Linux x86-64. All downloads are available from the same page.

As the oracle user account, extract the three packages you downloaded to a temporary directory. In this example, we will use /home/oracle/orainstall.

Extract the Oracle Clusterware package as follows:

# su - oracle
$ mkdir -p /home/oracle/orainstall
$ cd /home/oracle/orainstall
$ unzip 10201_clusterware_linux32.zip

$ cd /home/oracle/orainstall
$ unzip 10201_database_linux32.zip

$ cd /home/oracle/orainstall
$ unzip 10201_companion_linux32.zip

Perform the following checks on both Oracle RAC nodes in the cluster!

Before installing the Oracle Clusterware and Oracle RAC software, it is highly recommended to run the Cluster Verification Utility (CVU) to verify the hardware and software configuration. CVU is a command-line utility provided on the Oracle Clusterware installation media. It is responsible for performing various system checks to assist you with confirming the Oracle RAC nodes are properly configured for Oracle Clusterware and Oracle Real Application Clusters installation. The CVU only needs to be run from the node you will be performing the Oracle installations from (linux1 in this article). Note that the CVU is also run automatically at the end of the Oracle Clusterware installation as part of the Configuration Assistants process.

Prerequisites for Using Cluster Verification Utility

Install cvuqdisk RPM (Oracle Enterprise Linux and RHEL Users Only)

The first pre-requisite for running the CVU pertains to users running Oracle Enterprise Linux, Red Hat Linux, and SuSE. If you are using any of the above listed operating systems, then you must download and install the package cvuqdisk to both of the Oracle RAC nodes in the cluster. This means you will need to install the cvuqdisk RPM to both linux1 and linux2. Without cvuqdisk, CVU will be unable to discover shared disks and you will receive the error message "Package cvuqdisk not installed" when you run CVU.

The cvuqdisk RPM can be found on the Oracle Clusterware installation media in the rpm directory. For the purpose of this article, the Oracle Clusterware media was extracted to the /home/oracle/orainstall/clusterware directory on linux1. Note that before installing the cvuqdisk RPM, we need to set an environment variable named CVUQDISK_GRP to point to the group that will own the cvuqdisk utility. The default group is oinstall which is the group we are using for the oracle UNIX user account in this article.

Locate and copy the cvuqdisk RPM from linux1 to linux2 as the "oracle" user account:

$ ssh linux2 "mkdir -p /home/oracle/orainstall/clusterware/rpm"
$ scp /home/oracle/orainstall/clusterware/rpm/cvuqdisk-1.0.1-1.rpm linux2:/home/oracle/orainstall/clusterware/rpm

Perform the following steps as the "root" user account on both Oracle RAC nodes to install the cvuqdisk RPM:

$ su -
# cd /home/oracle/orainstall/clusterware/rpm
# CVUQDISK_GRP=oinstall; export CVUQDISK_GRP

# rpm -iv cvuqdisk-1.0.1-1.rpm
Preparing packages for installation...
cvuqdisk-1.0.1-1

# ls -l /usr/sbin/cvuqdisk
-rwsr-x--- 1 root oinstall 4168 Jun  2  2005 /usr/sbin/cvuqdisk

Verify Remote Access / User Equivalence

The CVU should be run from linux1 — the node we will be performing all of the Oracle installations from. Before running CVU, login as the oracle user account and verify remote access / user equivalence is configured to all nodes in the cluster. When using the secure shell method, user equivalence will need to be enabled for the terminal shell session before attempting to run the CVU. To enable user equivalence for the current terminal shell session, perform the following steps remembering to enter the pass phrase for each key that you generated when prompted:

# su - oracle
$ exec /usr/bin/ssh-agent $SHELL
$ /usr/bin/ssh-add
Enter passphrase for /home/oracle/.ssh/id_rsa: xxxxx
Identity added: /home/oracle/.ssh/id_rsa (/home/oracle/.ssh/id_rsa)

Verifying Oracle Clusterware Requirements with CVU

Once all prerequisites for running the CVU utility have been met, we can now check that all pre-installation tasks for Oracle Clusterware are completed by executing the following command as the "oracle" UNIX user account from linux1:

$ cd /home/oracle/orainstall/clusterware/cluvfy
$ mkdir -p jdk14
$ unzip jrepack.zip -d jdk14
$ CV_HOME=/home/oracle/orainstall/clusterware/cluvfy; export CV_HOME
$ CV_JDKHOME=/home/oracle/orainstall/clusterware/cluvfy/jdk14; export CV_JDKHOME
$ ./runcluvfy.sh stage -pre crsinst -n linux1,linux2 -verbose

Review the CVU report. Note that there are several errors you may ignore in this report.

If your system only has 1GB of RAM memory, you may receive an error during the "Total memory" check:

Check: Total memory
  Node Name     Available                 Required                  Comment
  ------------  ------------------------  ------------------------  ----------
  linux2        1009.65MB (1033880KB)     1GB (1048576KB)           failed
  linux1        1009.65MB (1033880KB)     1GB (1048576KB)           failed
Result: Total memory check failed.

As you can see from the output above, the requirement is for 1GB of memory (1048576 KB). Although your system may have 1GB of memory installed in each of the Oracle RAC nodes, the Linux kernel is calculating it to be 1033880 KB which comes out to be 14696 KB short. This can be considered close enough and safe to continue with the installation. As I mentioned earlier in this article, I highly recommend both Oracle RAC nodes have 2GB of RAM memory or higher for performance reasons.

The first error is with regards to finding a suitable set of interfaces for VIPs which can be safely ignored. This is a bug documented in Metalink Note 338924.1:

Suitable interfaces for the private interconnect on subnet "192.168.2.0":
linux2 eth1:192.168.2.101
linux1 eth1:192.168.2.100

ERROR:
Could not find a suitable set of interfaces for VIPs.

Result: Node connectivity check failed.

As documented in the note, this error can be safely ignored.

The last set of errors that can be ignored deal with specific RPM package versions that are not required with Oracle Enterprise Linux 5. For example:

• compat-db-4.0.14-5
• compat-gcc-7.3-2.96.128
• compat-gcc-c++-7.3-2.96.128
• compat-libstdc++-7.3-2.96.128
• compat-libstdc++-devel-7.3-2.96.128
• compat-libstdc++-devel-7.3-2.96.128

While these specific packages are listed as missing in the CVU report, please ensure that the correct versions of the compat-* packages are installed on both of the Oracle RAC nodes in the cluster. For example, in Oracle Enterprise Linux 4 Update 5, these would be:

• compat-gcc-32-3.2.3-47.3
• compat-gcc-32-c++-3.2.3-47.3
• compat-libstdc++-33-3.2.3-47.3

Checking the Hardware and Operating System Setup with CVU

The next CVU check to run will verify the hardware and operating system setup. Again, run the following as the "oracle" UNIX user account from linux1:

$ cd /home/oracle/orainstall/clusterware/cluvfy
$ ./runcluvfy.sh stage -post hwos -n linux1,linux2 -verbose

Review the CVU report. As with the previous check (pre-installation tasks for CRS), the check for finding a suitable set of interfaces for VIPs will fail and can be safely ignored

Also note you may receive warnings in the "Checking shared storage accessibility..." portion of the report:

Checking shared storage accessibility...

WARNING:
Unable to determine the sharedness of /dev/sde on nodes:
  linux2,linux2,linux2,linux2,linux2,linux1,linux1,linux1,linux1,linux1


Shared storage check failed on nodes "linux2,linux1".

If this occurs, this too can be safely ignored. While we know the disks are visible and shared from both of our Oracle RAC nodes in the cluster, the check itself may fail. Several reasons for this have been documented. The first came from Metalink indicating that cluvfy currently does not work with devices other than SCSI devices. This would include devices like EMC PowerPath and volume groups like those from Openfiler. At the time of this writing, no workaround exists other than to use manual methods for detecting shared devices. Another reason for this error was documented by Bane Radulovic at Oracle. His research shows that CVU calls smartclt on Linux, and the problem is that smartclt does not return the serial number from our iSCSI devices. For example, a check against /dev/sde shows:

# /usr/sbin/smartctl -i /dev/sde
smartctl version 5.33 [i686-redhat-linux-gnu] Copyright (C) 2002-4 Bruce Allen
Home page is http://smartmontools.sourceforge.net/

Device: Openfile Virtual disk     Version: 0
Serial number:
Device type: disk
Local Time is: Mon Sep  3 02:02:53 2007 EDT
Device supports SMART and is Disabled
Temperature Warning Disabled or Not Supported

At the time of this writing, it is unknown if the Openfiler developers have plans to fix this.