>>> Network Address for the Cluster Transport <<<

The private cluster transport uses a default network address of

172.16.0.0. But, if this network address is already in use elsewhere within your enterprise, you may need to select another address from the range of recommended private addresses (see RFC 1597 for details).

If you do select another network address, please bear in mind that

the Sun Clustering software requires that the rightmost two octets

always be zero.

The default netmask is 255.255.0.0; you may select another netmask,

as long as it minimally masks all bits given in the network address

and does not contain any holes.

Is it okay to accept the default network address [172.16.0.0] (y/n) y

8. Specify the netmask address of the private interconnect:

• To accept the default netmask address, press Y.
• To specify a different netmask address, press N, followed by the netmask address.

Is it okay to accept the default netmask [255.255.0.0] (y/n) y

Caution - You must press Y to accept the default netmask. Do not press Return. If you press Return, unpredictable behavior may occur.

9. Confirm the assigned names and addresses:

• If all are correct, press Y.
• If any of the values are not correct, press N, and the line number of the item to change, followed by the correct value.
• Press 99 to quit the update mode when all values are displayed correctly.

Management Server "MS-280R" (192.168.0.1) at 8:0:20:fd:ef:8 Option External Admin Network Setting ------- ------------------------------------------------------------- 1) Management Server's Default Router= 192.168.0.248 2) Terminal Server's Name= TC 3) Terminal Server's IP Address= 192.168.0.2 Option Cluster Environment and Public Network Settings ------ ------------------------------------------------------------- 4) Cluster Name= sc 5) Cluster Nodes Default Router= router_IP_address 6) Public Network Mask for the Cluster Nodes= 255.255.255.0 7) First Node's Name= node1 8) First Node's IP Address= node1_IP_address 9) First Node's Ethernet Address= 8:0:20:fb:29:8e 10) Second Node's Name= node2 11) Second Node's IP Address= node2_IP_addres 12) Second Node's Ethernet Address= 8:0:20:fb:1a:55 Option Internal Admin Network Settings Network Settings ------ ------------------------------------------------------------- 13) Internal Admin Network address= 10.0.0.0 14) Internal Admin Netmask= 255.255.255.0} 15) Admin IP Address for the Management Server= 10.0.0.1 16) Admin IP Address for the First Cluster Node= 10.0.0.2 17) Admin IP Address for the Second Cluster Node= 10.0.0.3 18) First T3's Host Name= T3-01 19) First T3's Admin Network IP Address= 10.0.0.4 20) First T3's Ethernet Address= 00:20:f2:00:3e:a6 21) Second T3's Host Name= T3-02 22) Second T3's Admin Network IP Address= 10.0.0.5 23) Second T3's Ethernet Address= 00:20:f2:00:04:d6 Option Cluster Private Interconnect Settings ------ ------------------------------------------------------------- 24) Private Interconnect Network Address= 172.16.0.0 25) Private Interconnect Netmask= 255.255.0.0 Option Finish Updates ------ ------------------------------------------------------------- 99) quit update mode....   Are all variables correctly set y/n? y

10. When prompted, choose one of the following volume manager products:

• For Solstice DiskSuite, press Y.
• For VERITAS Volume Manager, press N.

Storage Software Configuration <   The Cluster nodes can be configured with either *Solstice DiskSuite 4.2.1 or Veritas VM 3.1.1*   The default is Solstice DiskSuite 4.2.1. If the installer does not except the default, Veritas Volume Manager 3.1.1 will be installed. By choosing VXVM 3.1.1 you will be making a choose to encapsulate your boot disk.   A Veritas license is required for each node running VXVM 3.1.1. You will be prompted during the install of each cluster node to enter the license key.   Please have the license keys ready.   Do you want to install Solaris Volume Manager 4.2.1 (y/n) y

The configuration files for the management server are configured for root and swap mirroring.

================================================================ PLEASE WAIT: Setting up Management Server configuration files for jumpstart services Preparing Management Server for root and swap mirroring.   System will reboot after setup! ================================================================ Netra T1 200 (UltraSPARC-IIe 500MHz), No Keyboard OpenBoot 4.0, 1 GB memory installed, Serial #16641800. Ethernet address 8:0:20:fd:ef:8, Host ID: 80fdef08.

The management server reboots.

To Boot Up the T3 Arrays

---

Caution - During a recovery, do not power cycle the Sun StorEdge T3 arrays as mentioned in Step 1. Instead, proceed to Step 2.

---

1. Power on the Sun StorEdge T3 disk arrays.

============================================================= Power on T3's at this point. If T3's are already powered up, Power cycle them and let them reboot.   This will take 3-4 minutes!!   Please Wait until T3's have completely rebooted.   Press the Return key when T3's are ready! =============================================================

2. Press Return when the Sun StorEdge T3 disk arrays are finished booting.

To Set Up the Sun Management Center Software

Refer to the Sun Management Center 3.0 Installation Guide for additional details.

1. Choose to set up, or not set up the Sun Management Center (Sun MC) software:

• Press Y to set up Sun Management Center software (server, agent, and console) on the management server, and continue with Step 2.
• Press N to set up the management server without configuring the Sun Management Center software, and proceed to To Install the Oracle Software. You can install the Sun Management Center software at another time by running the Sun Management Center setup script
  (/opt/SUNWsymon/sbin/es-setup).

- Do you want to setup Sun MC 3.0 (y/n) y

---

Note - If you receive an error that reports Missing Product License, you can do one of the following: Ignore the error, and run the SunMC software as usual. The license is used for SunMC advanced monitoring features. The standard features do not require a license. Obtain a SunMC license. Visit: http://www.sun.com/solaris/sunmanagementcenter

---

2. Answer the Sun Management Center installation questions as shown in the following example:

----------------------------------- Sun Management Center Setup Program -----------------------------------   This program does setup of Sun Management Center components that are installed on your system.   Checking for Sun Management Center components installed on your system.   You have the following Sun Management Center components installed   Sun Management Center Server Sun Management Center Agent Sun Management Center Console   Sun Management Center successfully configured for java: "Solaris_JDK_1.2.2_08"   Configuring the system for setup, please wait.   This part of setup generates security keys used for communications between processes. A seed must be provided to initialize the keys. Please make sure you use the same seed for all the machines you install. You may like to keep record of this seed for future use.   Please enter the seed to generate keys:abc123 Please re-enter the seed to confirm:abc123   You should setup a user as a Sun Management Center administrator. This person will be added to the esadm and esdomadm groups. Please enter a user to be the Sun Management Center administrator:root   The Sun Management Center base URL is relative to the Sun Management Center Console. The Sun Management Center Console is able to request help documentation via the network.

If you have installed Sun Management Center help documentation in an http-accessible location within your network, you may specify this location. If Sun Management Center help is installed on the console host, simply accept the default value. Please enter base URL to Sun Management Center help [local]: <cr>   The base URL has been set to file:/opt/SUNWsymon/lib/locale ---------------------------------------------------------------- Starting Sun Management Center Service Availability Manager Setup ---------------------------------------------------------------- Setup for Service Availability Manager in progress, please wait. Setup of Service Availability Manager complete.

3. Specify whether you want to install the Sun Management Center Sun Fire 15K administration module.

• Press Y if you have Sun Fire 15K servers in your enterprise and you want to monitor them with Sun Management Center.
• Press N if you do not have Sun Fire 15K servers in your environment.

---------------------------------------------------------- Starting Sun Management Center Sun Fire 15000 Server Setup ----------------------------------------------------------   Would you like to setup this Sun Management Center package? [y|n|q] n

4. Specify whether you want to install the Sun Management Center Sun Fire 15K system controllers administration module.

• Press Y if you have Sun Fire 15K system controllers in your enterprise and you want to monitor them with Sun Management Center.
• Press N if you do not have Sun Fire 15K system controllers in your environment.

---------------------------------------------------- Starting Sun Management Center Sun Fire 15000 System Controller Server Setup ----------------------------------------------------   Would you like to setup this Sun Management Center package? [y|n|q] n

5. Specify whether you want to install the Sun Fire 6800/4810/4800/3800 administration module.

• Press Y if you have Sun Fire 6800/4810/4800/3800 servers in your enterprise and you want to monitor them. Answer the additional installation prompts.
• Press N if you do not have these Sun Fire servers in your environment.

For setting up Sun Fire (6800/4810/4800/3800) platform administration module you need to provide SC IP address, community strings, port numbers for domain agent etc.   Do you want to setup Sun Fire (6800/4810/4800/3800) platform administration module [y|n|q] n

6. Press Y to install the Sun Management Center Netra t administration module for monitoring the management server and Sun StorEdge T3 arrays.

---------------------------------------------------------- Starting Sun Management Center Netra t Setup ---------------------------------------------------------- ---> Platform Found: Netra T1 ---> Netra t add-on Agent package found! <--- Do you want to setup Netra t Config Reader for this platform? [y|n|q] y

7. Press Y to monitor the T3 arrays.

Do you want to setup T3 module [y|n|q] y

8. Press 3 to select the Add managed T3 routine.

Selecting 3 causes the script to prompt you with T3 setup questions.

---------------------------------- Sun Management Center StorEdge T3 Module Setup Program ----------------------------------   This program will configure the StorEdge T3, so that they can be managed by T3 Module   [1] List managed T3 [2] List available T3 [3] Add managed T3 [4] Remove managed T3 [5] Reconfigure managed T3 [6] Quit   Select the function you want to perform [1-6]: 3

9. Select 1 to display the list of available T3 arrays.

[1] add T3 from available T3 list. [2] add new T3 [3] return to the main menu Please press [1-3] 1

10. Type the line number of the first T3 array.

---

Note - Do not select any line numbers associated with the nodes.

---

Available T3: Name IP Address   1 node1 10.0.0.2 2 node2 10.0.0.3 3 T3-01 10.0.0.4 4 T3-02 10.0.0.5   Add available T3 [1-4] 3

11. Type the root password for the selected T3 array, and press Return.

---

Caution - During a recovery of the management server, the Sun StorEdge T3 arrays retain the passwords and IP addresses that you assigned during the initial installation. Resetting them during a recovery is not needed. If you are prompted for the array passwords and IP addresses, ignore the questions. The recovery process will continue after a time-out of a few seconds.

---

Input root password of T3-01:abc   Check SunMC token files... Check SunMC token files success. Check logging status... Check logging status success.   Press Enter to return: <CR>

12. Type 3 to configure Sun Management support for the second T3 array:

---------------------------------- Sun Management Center StorEdge T3 Module Setup Program ----------------------------------   This program will configure the StorEdge T3, so that they can be managed by T3 Module   [1] List managed T3 [2] List available T3 [3] Add managed T3 [4] Remove managed T3 [5] Reconfigure managed T3 [6] Quit   Select the function you want to perform [1-6]: 3

13. Type 1 to display the list of available arrays:

[1] add T3 from available T3 list. [2] add new T3 [3] return to the main menu Please press [1-3] 1

14. Type the line number of the second T3 array:

---

Note - Do not select any line numbers associated with the nodes.

---

Available T3: Name IP Address   1 node1 10.0.0.2 2 node2 10.0.0.3 3 T3-02 10.0.0.5   Add available T3 [1-3] 3

15. Type the root password for this T3 array, and press Return.

---

Caution - During a recovery of the management server, the Sun StorEdge T3 arrays retain the passwords and IP addresses that you assigned during the initial installation. Resetting them during a recovery is not needed. If you are prompted for the array passwords and IP addresses, ignore the questions. The recovery process will continue after a time-out of a few seconds.

---

Input root password of T3-02:abc   Check SunMC token files... Check SunMC token files success. Check logging status... Check logging status success.   Press ENTER to return: <CR>

16. Type 6 to exit the T3 module setup program.

The Sun Management Center software will install the module that supports monitoring of T3 arrays.

---------------------------------- Sun Management Center StorEdge T3 Module Setup Program ----------------------------------   This program will configure the StorEdge T3, so that they can be managed by T3 Module   [1] List managed T3 [2] List available T3 [3] Add managed T3 [4] Remove managed T3 [5] Reconfigure managed T3 [6] Quit   Select the function you want to perform [1-6]: 6

17. Specify whether you want to install the Sun Management Center CP2000 administration module.

• Press Y to use Sun Management Center to monitor any CP2000 systems (SPARCengine based systems with compact PCI) in your enterprise network and you want to monitor them with Sun Management Center.
• Press N if you do not have CP2000 systems in your environment.

---------------------------------------------------------- Starting Sun Management Center CP2000 Application Setup ---------------------------------------------------------- PLATFORM_FILE= /var/opt/SUNWsymon/platform/platform.prop /var/opt/SUNWsymon/platform ---> Visa application add-on Agent package found! <--- Do you want to setup the CP2000 applications for this platform? [y|n|q] n

18. If you receive the following question, specify whether you want to install the Sun Management Center CP2000/CP1500 server module.

• Press Y to use Sun Management Center to monitor any CP2000/CP1500 platforms in your enterprise and you want to monitor them with Sun Management Center.
• Press N if you do not have CP2000/CP1500 platforms in your environment.

----------------------------------------------------------- Starting Sun Management Center CP2000/CP1500 Server Setup ----------------------------------------------------------- PLATFORM_FILE= /var/opt/SUNWsymon/platform/platform.prop /var/opt/SUNWsymon/platform ---> Visa Server add-on package found! <--- Do you want to setup the Server Package for the CP2000/CP1500 platforms? [y|n|q] n

19. Press N to skip starting the Sun MC components.

---

Caution - You should press N to the following prompt. Pressing Y causes a list of exception errors that interfere with the script output. If you press Y, wait until the error messages complete, then press Return to be prompted to install the Oracle software, as in the next step. After the nodes have completed their configuration, you can manually start the Sun MC server and agent components on the management server. Alternatively, you can wait until the management server is rebooted, at which time the server and agent will be started automatically.

---

---------------------------------------------------------- Starting Sun Management Center WGS Setup ---------------------------------------------------------- Using /var/opt/SUNWsymon/cfg/console-tools.cfg /var/opt/SUNWsymon/cfg/tools-extension-j.x updated.   Sun Management Center setup complete. Please wait, Sun Management Center database setup in progress. It may take 15 to 20 minutes   Do you want to start Sun Management Center agent and server components now [y|n|q] n

To Install the Oracle Software

1. Press Y to install the Oracle9i with Sun Cluster High Availability software.

Do you want to install Oracle9i HA (y/n) y Do you want to install Oracle9i and SVM 4.2.1 (y/n) n Oracle9i HA and Volume Manager 3.1.1 will be installed.

The Oracle software license will be printed.

2. When prompted, press Return to review the Oracle license terms.

3. Press Y to agree to the Oracle license terms.

The login prompt will appear.

To Install the Clustered Node Software

---

Note - When executing the ccp(1M) command remotely, you must ensure that the DISPLAY shell environment variable is set to the IP address of the local host.

---

1. Type the command /user/openwin/bin/xhost 192.168.0.1 (the administration IP address of the management server) to enable your windows manager to display screens from the management server.

2. Log in to the management server as the superuser (root).

3. Set and export the DISPLAY environment variable to the IP address of the local host:

# DISPLAY=local_host_IP_address:0.0; export DISPLAY

4. Start the Cluster Control Panel:

# ccp $CLUSTER &

This command displays the Cluster Control Panel (FIGURE 3-3).

 FIGURE 3-3 Cluster Control Panel

5. In the Cluster Control Panel window, double-click on the cconsole icon to display a Cluster Console window for each cluster node (FIGURE 3-4).

 FIGURE 3-4 Cluster Node Console Windows

To type text into both node windows simultaneously, click the cursor in the Cluster Console window and type the text. The text does not display in the Cluster Console window. This ensures that both nodes execute the commands simultaneously.

---

Note - The console windows for both clustered nodes are grouped (the three windows move in unison). To ungroup the Cluster Console window from the node windows, select Options from the Hosts menu, and deselect the Group Term Windows checkbox.

---

6. Use the ccp(1M) Cluster Console window to type the following command into both nodes simultaneously:

boot net - install

---

boot net - You must use spaces before and after the dash (-) character in the "install" command string.

---

The following software is automatically installed on both cluster nodes:

• Solaris 8 10/01 operating environment
• VERITAS Volume Manager 3.1.1
• Sun Cluster 3.0 7/01 and recommended patches
• Oracle9i Database Edition, version 9.0.1.2

You may see the following error:

Boot device: /pci@8,700000/network@5,1: File and args: Timeout waiting for ARP/RARP packet Timeout waiting for ARP/RARP packet Timeout waiting for ARP/RARP packet . . .

If so, it is likely that you need to correct one of the following:

• Naming service conflict--A naming service was selected during the configuration of the management server.
• No network route--For network environments with a production and an administration network, the two network backbones must be routed so that the clustered nodes can jumpstart from the management server. If this is the case, you need to temporarily provide two Ethernet cables and connect the eri0 network interfaces on both clustered nodes to the supplied ethernet hub in the Clustered Platform. Once the installation and setup is complete, remove the temporary connections from the ethernet hub and the nodes, and reconnect the cables from the clustered nodes eri0 to the production network.

---

Note - Press Ctrl+] and send brk to stop the Timeout waiting for ARP/RARP packet errors.

---

7. Take one of the following actions based on the volume manager product you chose:

• Select Solstice DiskSuite--continue with Step 8 below.
• Select VERITAS Volume Manager--proceed to To Configure VERITAS Volume Manager.

8. Log in to each clustered node as a superuser (password is abc) and change the default password to a secure password:

# passwd passwd: Changing password for root New password: secure-password-choice Re-enter new password: secure-password-choice

9. Install the Sun Management Center agent on the clustered nodes.

Installing the Sun Management Center agent enables Sun Management Center to monitor the nodes. For installation instructions, refer to the Sun Management Center 3.0 Installation Guide.

10. Configure the Sun StorEdge T3 array shared disk storage.

Configure the T3 array volume configuration if you do not plan to use the default T3 configuration. Refer to the T3 documentation that shipped with your Clustered Platform.

Configure the storage using Solstice DiskSuite, then create disk sets, meta devices, and file systems that suit your needs. Refer to the Solstice DiskSuite documentation.

11. Proceed to To Run the HAOracle.init Script.

To Configure VERITAS Volume Manager

---

Caution - The following instructions apply only to configurations that are using VERITAS Volume Manager instead of Solstice DiskSuite. Only follow this procedure only if you selected to use VERITAS Volume Manager.

---

---

Note - You will not see these questions if you chose to install Solstice DiskSuite.

---

You must have two valid VERITAS license keys (one for each node) to complete this procedure.

1. In the Cluster Console window (cconsole) choose to encapsulate the boot disk.

Do you want Volume Manager to encapsulate root [yes]? <cr>   Where is the Volume Manager cdrom? /net/ManagementServer_IP_address/jumpstart/Packages/VM3.1.1

---

Note - Ignore the references to "cdrom" in the computer output.

---

2. In each individual node window, enter the license key for that node (obtained from VERITAS) to register your version of VERITAS Volume Manager:

---

Note - You must enter your own valid VERITAS license keys for the script to complete successfully. A VERITAS license key is made up of 23 digits and is formatted as follows: 1234 5678 1234 5678 1234 567 (This is an example only. Do not use this number.)

---

---

Caution - You must enter valid license keys. If the script detects an error in the license key, the script will not complete successfully.

---

(For node 1) Volume Manager installation is complete. Oct 8 15:34:03 node1 vxdmp: NOTICE: vxvm:vxdmp: added disk array 60020f20000004d60000000000000000 Oct 8 15:34:03 node1 vxdmp: NOTICE: vxvm:vxdmp: added disk array 60020f2000003ea00000000000000000   Please enter a Volume Manager license key: node1_VERITAS_license_key

 

(For node 2) Volume Manager installation is complete. Oct 8 15:34:03 node2 vxdmp: NOTICE: vxvm:vxdmp: added disk array 60020f20000004d60000000000000000 Oct 8 15:34:03 node2 vxdmp: NOTICE: vxvm:vxdmp: added disk array 60020f2000003ea00000000000000000   Please enter a Volume Manager license key: node2_VERITAS_license_key

3. In the Cluster Console window, wait for the installer script to display the login prompt:

================================================================ Setting up root mirror. This will take approximately 45min's   Please wait .... ================================================================ The system is ready. node1 console login:

4. In each individual node window, log in to each clustered node as a superuser (password is abc) and change the default password to a secure password:

node1 console login: root Password: abc # passwd passwd: Changing password for root New password: secure-password-choice Re-enter new password: secure-password-choice

 

node2 console login: root Password: abc # passwd passwd: Changing password for root New password: secure-password-choice Re-enter new password: secure-password-choice

5. Install the Sun Management Center agent on the clustered nodes.

Installing the Sun Management Center agent enables Sun Management Center to monitor the nodes. For installation instructions, refer to the Sun Management Center 3.0 Installation Guide.

6. Configure the Sun StorEdge T3 array shared disk storage.

Configure the storage using VERITAS Volume Manager. Configure the volumes and file systems to suit your needs. Refer to the VERITAS Volume Manager documentation.

---

Note - The VERITAS Volume Manager VMSA packages are available on the management server in the /jumpstart/Packages/VM3.1.1 directory. If you want to use the VERITAS GUI, install the VMSA packages with the pkgadd command. Once installed, the Volume Manager man pages will be placed in the /opt/VRTS/man directory. Add this directory to your MANPATH variable to access these pages with the man command.

---

7. Proceed to To Run the HAOracle.init Script.

To Run the HAOracle.init Script

This script prompts you for database configuration information. The information is saved in the HAOracle.env file and used later by the HAOracleConfig.init script. At the end of this script, you have the chance to verify the configuration information and make changes.

Perform the following steps on node 1, as listed in the /etc/inet/hosts file.

1. As superuser on one of the nodes, start the HAOracle.init script:

# /etc/opt/SUNWords/HAOracle.init

---

Note - When answering these script questions, a default value is assigned if you press Return without specifying any value. The default values are acceptable values for many installations. Therefore, if you are uncertain about a response to a question, press Return to accept the default value. You could press Return to most of the questions (unless marked mandatory) and end up with a usable Oracle9i HA database configuration.

---

2. Accept the Oracle license for all nodes:

If you agree to the license terms above and are in compliance with each of the export restrictions listed above, please click on the "I accept the License Terms and Export Restrictions" button below. Otherwise please click "Exit" to end the installation of the Software.   Do you agree to the terms of this license for all nodes? [default=y] [y,n,?] Y

3. Press Return, or enter a value at each question mark.

In this example, and many that follow, the defaults are selected.

---

Note - Do not create Oracle database files on the local disk of a node. The mount point for shared storage should always be /global/mount_point.

---

Please enter Oracle shared storage mount point [default: /global/oradata] [?] <cr>   Thank you. You will now be asked for a series of directories in which to store the various Oracle data files. Be sure to make them all subordinate to the mount point you have chosen for shared storage, /global/oradata.   Please enter directory for control file one [default: /global/oradata/orcl/Cntl_file_1] [?] <cr>   Please enter directory for control file two [default: /global/oradata/orcl/Cntl_file_2] [?] <cr>   Please enter directory for control file three [default: /global/oradata/orcl/Cntl_file_3] [?] <cr>   Please enter directory for redo log file one [default: /global/oradata/orcl/Log_group_1] [?] <cr>   Please enter directory for redo log file two [default: /global/oradata/orcl/Log_group_2] [?] <cr>   Please enter directory for redo log file three [default: /global/oradata/orcl/Log_group_3] [?] <cr>   Please enter directory for archived redo log files [default: /global/oradata/orcl/Arch] [?] <cr>   Please enter directory for Oracle data files [default: /global/oradata/orcl/Data_files] [?] <cr>   Please enter directory for alert log and trace files [default: /global/oradata/admin/orcl] [?] <cr>

4. Enter the Oracle user login name and password:

Please enter database login name [doctor]: <cr> Please enter database password [heartbeat]: <cr>

5. Enter the total database size in Gbytes:

Please enter total database size on disk in GB [10]: <cr>

The minimum database size limits differ based on the type of database you are planning to use:

• 3 Gbytes for Decision Support System (DSS)
• 2 Gbytes for On-line Transaction Processing (OLTP)

The maximum database size limit is based on the amount of storage space on your nodes.

6. Enter the following Oracle names:

Please enter failover resource group name [oracle-rg]: <cr> Please enter server application resource name [oracle-server]: <cr> Please enter listener application resource name [oracle-listener]: <cr> Please enter HA storage resource name [HAStorage_res]: <cr>

7. Select either VERITAS Volume Manager (Y) or Solstice DiskSuite (N):

Your answer should correspond with your choice during the management server installation.

• If you press N, you will see:

Use Veritas Volume Manager? [default: Y] [y,n,?] n

• If you press N, you will be asked for the VERITAS group name:

Use Veritas Volume Manager? [default: Y] [y,n,?] y Please enter the Veritas group name [oradata-dg]: <cr>

8. Choose to set up either a DSS or OLTP database.

Press N for an on-line transaction processing (OLTP) database. Press Y for a decision support system (DSS) database.

Set up DSS server? [default: N] [y,n,?] y

9. Enter the Oracle logical host name and IP address:

Replace the examples shown with the host name and IP address appropriate for your site:

Please enter logical host name: hostname Please enter logical IP address: IP_address

10. Define the following Oracle database parameters (or press Return to apply the default value).

The following table lists the defaults for these parameters:

Parameter	Default Value
fault probe number of cycles	5 tries
fault probe time-out value	60 seconds
user environment file	No default
LOG_BUFFER	524288 bytes
DB_CACHE_SIZE	67108864 bytes
SHARED_POOL_SIZE	67108864 bytes
SORT_AREA_SIZE	DSS- 1048576 bytes, OLTP- 524288 bytes
JAVA_POOL_SIZE	67108864 bytes

Please enter fault probe number of cycles [optional]: <cr> Please enter fault probe time-out value [optional]: <cr> Please enter Oracle user environment file [optional]: <cr> Please enter Oracle parameter LOG_BUFFER in bytes [optional]: <cr> Please enter Oracle parameter DB_CACHE_SIZE in bytes [optional]: <cr> Please enter Oracle parameter SHARED_POOL_SIZE in bytes [optional]: <cr> Please enter Oracle parameter SORT_AREA_SIZE in bytes [optional]: <cr> Please enter Oracle parameter JAVA_POOL_SIZE in bytes [optional]: <cr>

This completes the questions. The values are displayed.

11. Review the values assigned to each parameter.

Now you will see the configuration options as you have chosen them. If you would like to modify a line, note its line number.   Press any key to continue... 1 ORADATA_MT_PT=/global/oradata 2 CNTL_FILE_1=/global/oradata/orcl/Cntl_file_1 3 CNTL_FILE_2=/global/oradata/orcl/Cntl_file_2 4 CNTL_FILE_3=/global/oradata/orcl/Cntl_file_3 5 LOG_FILE_1=/global/oradata/orcl/Log_group_1 6 LOG_FILE_2=/global/oradata/orcl/Log_group_2 7 LOG_FILE_3=/global/oradata/orcl/Log_group_3 8 ARCHIVED_LOG_FILES=/global/oradata/orcl/Arch 9 DATA_FILE=/global/oradata/orcl/Data_files 10 TRACE_FILE=/global/oradata/admin/orcl 11 DB_LOGIN=doctor 12 DB_PASS=heartbeat 13 ORADATA_SZ_GB=10 14 FO_RES_GROUP_NM=oracle-rg 15 SERVER_RES_NM=oracle-server 16 LISTENER_RES_NM=oracle-listener 17 HA_STORAGE_RES_NM=HAStorage_res 18 USE_VERITAS_VM=N 19 SIXTY_FOUR_BIT=N 20 DSS_SERVER=Y 21 LOGICAL_HOST_NAME=HAOracle 22 LOGICAL_IP_ADDR=129.153.47.202 Note number of line to change. Then press "q". Enter line number to edit. Else press Return to save.

---

Note - Do not attempt to change the value for SIXTY_FOUR_BIT=N. Only the 32-bit version of Oracle9i software is available.

---

12. If all the values are correct, press Return, otherwise continue to Step 13.

After you press Return, the assigned values are written to the HAOracle.env file in the current working directory.

Continue to To Run the HAOracleConfig.init Script.

13. If you want to change a value, enter the corresponding line number.

Example of making a change:

Note number of line to change. Then press "q". q Enter line number to edit. Else press Return to save. 13   Please enter new value for ORADATA_SZ_GB: 12   Now you will see the configuration options as you have chosen them. If you would like to modify a line, note its line number.   Press any key to continue... 1 ORADATA_MT_PT=/global/oradata 2 CNTL_FILE_1=/global/oradata/orcl/Cntl_file_1 3 CNTL_FILE_2=/global/oradata/orcl/Cntl_file_2 4 CNTL_FILE_3=/global/oradata/orcl/Cntl_file_3 5 LOG_FILE_1=/global/oradata/orcl/Log_group_1 6 LOG_FILE_2=/global/oradata/orcl/Log_group_2 7 LOG_FILE_3=/global/oradata/orcl/Log_group_3 8 ARCHIVED_LOG_FILES=/global/oradata/orcl/Arch 9 DATA_FILE=/global/oradata/orcl/Data_files 10 TRACE_FILE=/global/oradata/admin/orcl 11 DB_LOGIN=doctor 12 DB_PASS=heartbeat 13 ORADATA_SZ_GB=12 14 FO_RES_GROUP_NM=oracle-rg 15 SERVER_RES_NM=oracle-server 16 LISTENER_RES_NM=oracle-listener 17 HA_STORAGE_RES_NM=HAStorage_res 18 USE_VERITAS_VM=N 19 SIXTY_FOUR_BIT=N 20 DSS_SERVER=Y 21 LOGICAL_HOST_NAME=HAOracle 22 LOGICAL_IP_ADDR=129.153.47.202 Note number of line to change. Then press "q". Enter line number to edit. Else press Return to save. <cr> HAOracle.init: Final configuration written to HAOracle.env.   Press any key to continue...

14. After all of the values are correct, press Return.

The HAOracle.init script is complete and the values are written to the HAOracle.env file in the current working directory. Continue to To Run the HAOracleConfig.init Script.

To Run the HAOracleConfig.init Script

1. As superuser on the clustered node, start the HAOracleConfig.init script:

# /etc/opt/SUNWords/HAOracleConfig.init

The script runs for several minutes and there are no questions to answer. It uses the values in the HAOracle.env file (which was populated by the HAOracle.init script). By default, the HAOracleConfig.init script looks for the HAOracle.env file in the current working directory. If you change directories between script execution, you must specify the full pathname to the HAOracle.env file as an argument to the HAOracleConfig.init script.

Refer to Appendix I for sample output from this script.

---

Caution - Do not interrupt this script by pressing Ctrl-C, or any interruption method, or else the ability to determine the state of the database configuration will be unpredictable. If errors occur, let the script finish, and continue with these instructions.

---

2. Check the output for successful completion.

When the script completes you should see the following:

. . . HAOracleConfig.init: Bringing resource group oracle-rg online. HAOracleConfig.init: completed successfully. #

After successful completion, an Oracle database named ORCL is started on the node on which you ran this script (this becomes the primary node). If no errors were reported, and the script completed successfully, continue to To Verify Basic Cluster Functionality.

If errors were reported, continue to To Recover From Errors That Occurred During Script Execution.

To Recover From Errors That Occurred During Script Execution

Errors might occur when values assigned during the HAOracle.init script conflict with the hardware configuration of the system. The error messages attempt to explain the nature of the error. Refer to the online Operation Guide for error message documentation (see Clustered Database Platform Documentation)

---

Caution - You only need to perform this procedure if you encountered an error while running the HAOracleConfig.init script. Do not run the HACleanup script if no errors occurred.

---

The following steps must be performed on the node for which you ran the two HAOracle scripts.

1. When the HAOracleConfig.init script is finished, run the HACleanup script:

# /etc/opt/SUNWords/HACleanup

After you confirm that you want to continue, the script runs for a few minutes and there are no questions to answer. The script unconfigures any cluster objects which might have been created during the previous run of the HAOracleConfig.init script. Refer to Appendix I to see an example of the output of this script.

2. Edit the HAOracle.env file that was created by the HAOracle.init script.

The HAOracle.env file contains the configuration information based on the answers you supplied to the HAOracle.init script. The HAOracle.env file is located in the current working directory from which you ran the HAOracle.init script.

Change any parameter value that caused the error to occur. Save your changes.

---

Note - An alternate method to editing the HAOracle.env file is to rerun the HAOracle.init script as described in To Run the HAOracle.init Script.

---

3. Rerun the HAOracleConfig.init script.

See To Run the HAOracleConfig.init Script.

To Verify Basic Cluster Functionality

1. Perform the following command to check the active nodes:

# scstat -p

The status of the nodes is displayed.

2. As superuser, run the scswitch command to test switching from the primary node to the other node:

# scswitch -S -h other_node_name

---

Note - The initial database is configured with the default Oracle passwords. For the user sys, the password is change_on_install. For the user system, the password is manager. Change these two passwords by using the alter user command within SQL*Plus.

---

3. Proceed to Finishing Up.

---

Finishing Up

You need to perform additional steps depending on your environment. This section contains references to additional steps that you might need to perform.

To Finish the Installation

1. Install and configure your highly available application(s) for the clustered environment.

For additional details, refer to the manufacturers' documentation that accompanied your highly available application(s).

Refer to the following appendixes for additional information about configuring the following software that came with your Clustered Platform:

• Appendix E, Installing Sun StorEdge Component Manager 2.2 on the Management Server
• Appendix F, Configuring the Sun Ray Server Software on the Management Server
• Appendix G, Setting Up the NTP Server on the Management Server

2. Establish network automatic failover (NAFO), resource groups, logical hosts, and data services to enable your application(s) under the Sun Cluster 3.0 infrastructure. Refer to the Sun Cluster 3.0 documentation. The path to the Sun Cluster data services is:

/net/sc3sconf1-ms/jumpstart/Packages/SC3.0u1/scdataservices_3_0_u2

NAFO configuration and activation information is in Appendix H.

---

Note - Before you use an editor in a Cluster Console window, verify that the TERM shell environment value is set and exported to a value that emulates the kind of input terminal your are using on the local system.

---

---

Note - You can stop the clustered nodes and access the OpenBoot PROM prompt by positioning the cursor in the Cluster Console window and pressing Ctrl+]. This control character sequence displays the telnet prompt. Typing send brk (equivalent to a Stop-A) at the telnet prompt forces access to the OpenBoot PROM prompt.

---

3. Optionally, enable the Oracle GUI configuration assistant to access the nodes.

Certain Oracle GUI configuration assistants such as dbca and netca require the ability to issue a remote shell (rsh) to the nodes. For this to take place, a .rhosts file must reside in the home directory of each node (for security reasons, the default configuration does not create a .rhosts file). The .rhosts file for each node is populated with the node name of each node, and private interconnect IP addresses (for the local node), followed by the oracle UNIX user name. For security, delete both .rhosts files when you are finished running dbca or netca.

Content for a .rhosts file:

cluster_node1_name oracle cluster_node2_name oracle node_private_interconnect_IP_1 oracle node_private_interconnect_IP_2 oracle

---

Note - You can run /usr/sbin/ifconfig -a on a node to determine the private interconnect IP addresses.

---

4. Place the Oracle database into archive log mode.

Choose a location where the Oracle redo log files are copied, and specify the location in the server parameter file. An area on the shared storage device can be used for this purpose. This location should be specified on a UFS file system, not on a raw partition. For more information on these topics, refer to the Oracle documentation.

5. Determine and set up your backup strategy and methods.

For backing-up the Oracle database, if you chose not to place the database into archive log mode, then only offline backups can be performed. In this case, you need to shutdown the instance during the backup, and recovery is only available back to the time of the last offline backup. For more information on these topics, refer to the Oracle documentation.

---

After the Installation

This section contains information about what has been installed and configured on the cluster nodes and what you can do with the software.

Post-Installation Modification of the /etc/hosts File

After the management server finishes installing the software on the nodes, you must update the /etc/hosts file so that it corresponds to the cluster platform network configuration.

To Update the /etc/hosts File

1. On the management server, use a text editor to open the /etc/hosts file.

2. Delete the first two occurrences of the -admin text string (shown highlighted in the following example):

# Physical Hosts (Physical Addresses) 129.153.47.181 test #Management Server 129.153.47.71 TC #Terminal Concentrator 129.153.47.120 node1-admin #First Cluster Node 129.153.47.121 node2-admin #Second Cluster Node 10.0.0.1 test-admin #Admin Network 10.0.0.2 node1 #First Cluster Node admin network 10.0.0.3 node2 #Second Cluster Node admin network 10.0.0.4 T3-01 #First T3 Host Name 10.0.0.5 T3-02 #Second T3 Host Name

3. Append -admin to the two internal administration node names (shown highlighted in the following example):

# Physical Hosts (Physical Addresses) 129.153.47.181 test #Management Server 129.153.47.71 TC #Terminal Concentrator 129.153.47.120 node1 #First Cluster Node 129.153.47.121 node2 #Second Cluster Node 10.0.0.1 test-admin #Admin Network 10.0.0.2 node1-admin #First Cluster Node admin network 10.0.0.3 node2-admin #Second Cluster Node admin network 10.0.0.4 T3-01 #First T3 Host Name 10.0.0.5 T3-02 #Second T3 Host Name

4. Save your changes and quit the editing session.

Post-Installation Removal of the /.rhosts File

For security reasons, when the installation and setup of the cluster platform is complete, remove the root user /.rhosts file on each cluster node to prevent unauthorized access to the nodes. This file is not typically needed after the cluster installation is complete. However, some cluster agents may require the root user to have remote access to the cluster nodes. Refer to the agent documentation for more details.

Database Recovery

You should place each starter database provide with the Clustered Database Platform into archive log mode. You must choose a location to which the redo log files may be copied and to specify it in the server parameter file. For Oracle9i with Sun Cluster High Availability software, you can use an area on the shared storage device; however, the location must be on a UFS file system, not a raw partition.

Starter Database

The starter database is configured with the default Oracle passwords. For the sys user, the password is change_on_install. For the system user, the password is manager. You should change these passwords by using the alter user command in SQL*Plus.

Backup Method

You should choose a backup method to be performed on a regular basis, such as offline or online. You must provide the scripts. If you choose not to place the database into archive log mode, then only offline backups can be performed. An offline backup involves a shutdown of the instance, which means that the recovered data would only reflect the last offline backup. For more information on these topics or to read about Oracle Recovery Manager, refer to your Oracle software documentation.

Operation Instructions

You can access operation instructions in the Clustered Database Platform 280/3 system documentation and the Oracle documentation (see Clustered Database Platform Documentation).

Clustered Database Platform 280/3 With Oracle9i and Sun Cluster HA Software Installation Guide

816-4071-10

Copyright © 2002, Sun Microsystems, Inc. All rights reserved.