7 Oracle Software Cloning Using Oracle Universal Installer

7.4.2.1 Viewing Log and Error Files

After cloning, you can view the status of the clone operation by navigating to the $ORACLE_HOME/clone/log directory and by reviewing the *.log and *.err files. For more information, refer to the section "Locating and Viewing Log Files".

7.5.2.1 Viewing Log and Error Files

After cloning, you can view the status of the clone operation by navigating to the $ORACLE_HOME/clone/log directory and by reviewing the *.log and *.err files. For more information, refer to the section "Locating and Viewing Log Files".

7.6.1.1 Cloning Oracle Clusterware on UNIX and Linux System-Based Environments

Complete the following steps to clone Oracle Clusterware on UNIX and Linux systems:

1. If you do not have a shared Oracle Clusterware home, then tar the Oracle Clusterware home from an existing node and copy it to the new node. Use $CRS_HOME as the destination Oracle Clusterware home on the new node. If you have a shared Oracle Clusterware home, then skip this step.

   Note:

   For more information on archiving, refer to section "Source Preparation Phase".

2. Unarchive the home on the new nodes. In the case of shared homes, unarchive the home only once on the nodes.

   Note:

   For more information on archiving, refer to section "Cloning Phase".

3. If you do not have a shared Oracle Clusterware home, then on the new node, navigate to the $ORACLE_HOME/clone/bin directory and run the following command where Oracle_home_name is the name of the Oracle home, new_node is the name of the new node, new_node-priv is the private interconnect protocol address of the new node, new_node-vip is the virtual interconnect protocol address of the new node, and central_inventory_location is the location of the Oracle central inventory:

   perl clone.pl ORACLE_HOME=<$CRS_HOME> ORACLE_HOME_NAME=Oracle_home_name '-On_storageTypeVDSK=2' '-On_storageTypeOCR=2' '-O"sl_tableList={new_node:new_node-priv:new_node-vip}"' '-O-noConfig''-O"INVENTORY_LOCATION=central_inventory_location"'
   
   

   If you have a shared Oracle Clusterware home, then append the -cfs option to the command example in this step and provide a complete path location for the cluster file system. Ensure that n_storageTypeOCR and n_storageTypeVDSK has been set to 2 for redundant storage. Ensure that this value is set to 1 for non-redundant storage. In this case, the mirror locations will also have to be specified.

4. On the new node, go to the directory that contains the central Oracle inventory. Run the orainstRoot.sh script to populate the file /etc/oraInst.loc with information about the central inventory location. On the new node, go to the $CRS_HOME directory and run ./root.sh. This starts the Oracle Clusterware on the new node.

5. Determine the remote port to use in the next step by running the following command from the $CRS_HOME/opmn/conf directory:

   cat ons.config | grep remoteport
   
   

6. On the new node, run the following from $CRS_HOME/bin directory where racgons is the Oracle RAC Notification Service Utility, new_node is the name of the new node, and remote_port is the value from the output of the previous step:

   ./racgons add_config new_node:<Remote_Port>
   
   

7. Execute the following command to get the interconnect information. You can use this information in the next step.

   $CRS_HOME/bin/oifcfg iflist –p
   
   

8. Execute oifcfg command as follows:

   oifcfg setif -global <interface_name>/<subnet>:public <inteface_
   name>/<subnet>:cluster_interconnect [<interface_name>/<subnet>:public
    <inteface_name>/<subnet>:cluster_interconnect .......]
   

7.6.1.2 Cloning Oracle RAC Software on UNIX and Linux System-Based Environments

Complete the following steps to clone Oracle Database with RAC software on UNIX and Linux systems:

1. If you do not have a shared Oracle Database home, then tar the Oracle RAC home from the existing node and copy it to the new node. Assume that the location of the destination Oracle RAC home on the new node is $ORACLE_HOME. Otherwise, skip this step.

   Note:

   For more information on archiving, refer to section "Source Preparation Phase".

2. Unarchive the home on the new nodes. In the case of shared homes, unarchive the home only once on the nodes.

   Note:

   For more information on unarchiving, refer to section "Cloning Phase".

3. On the new nodes, go to the $CRS_HOME/oui/bin directory and run the following command where new_node2 and new_node3 are the names of the new nodes, and Oracle_home_name is the name of the Oracle home:

   perl clone.pl ORACLE_HOME=<Path to the Oracle_Home being cloned> ORACLE_HOME
   _NAME=<Oracle_Home_Name for the Oracle_Home being cloned> '-O"CLUSTER_
   NODES={new_node_2,new_node_3}"' '-O"LOCAL_NODE=new_node_2"'
   
   

   If you have a shared Oracle Database home, then append the -cfs option to the command example in this step and provide a complete path location for the cluster file system

   Note:

   Set the LOCAL_NODE to the node on which you run the clone command.

4. On the new node, go to the $ORACLE_HOME directory and run the following command:

   ./root.sh
   
   

5. On the new node, run Net Configuration Assistant (NETCA) to add a listener.

6. From the node that you cloned, run Database Configuration Assistant (DBCA) to add the new instance.

7.6.2.1 Cloning Oracle Clusterware on Windows System-Based Environments

Complete the following steps to clone Oracle Clusterware on Windows system computers:

1. If you do not have a shared Oracle Clusterware home, then zip the Oracle Clusterware home from the existing node and copy it to the new node. Unzip the home on the new node in the equivalent directory structure as the directory structure in which the Oracle Clusterware home resided on the existing node. For example, assume that the location of the destination Oracle Clusterware home on the new node is %CRS_HOME%. Skip this step if you have a shared Oracle Clusterware home.

   Note:

   For more information on zipping and unzipping, refer to sections "Source Preparation Phase" and "Cloning Phase".

2. If you do not have a shared Oracle Clusterware home, then on the new node, navigate to the $ORACLE_HOME/clone/bin directory and run the following command where Oracle_home_name is the name of the Oracle home, new_node is the name of the new node, new_node-priv is the private interconnect protocol address of the new node, new_node-vip is the virtual interconnect protocol address of the new node, and central_inventory_location is the location of the Oracle central inventory:

   perl clone.pl ORACLE_HOME=<CRS_HOME> ORACLE_HOME_NAME=<CRS_HOME_NAME> '-On_storageTypeVDSK=2' '-On_storageTypeOCR=2' '-O"sl_tableList={node2:node2-priv:node2-vip, node3:node3-priv:node3-vip}"' '-O"ret_PrivIntrList=<private interconnect list>"' '-O"sl_OHPartitionsAndSpace_valueFromDlg={partition and space information}"' '-O-noConfig'
   
   

   If you have a shared Oracle Clusterware home, then append the -cfs option to the command example in this step and provide a complete path location for the cluster file system. Ensure that n_storageTypeOCR and n_storageTypeVDSK has been set to 2 for redundant storage. Ensure that this value is set to 1 for non-redundant storage. In this case, the mirror locations will also have to be specified. On the other nodes, execute the same command, by passing an additional argument PERFORM_PARTITION_TASKS=FALSE.

   perl clone.pl ORACLE_HOME=<CRS_HOME> ORACLE_HOME_NAME=<CRS_HOME_NAME> '-On_
   storageTypeVDSK=2' '-On_storageTypeOCR=2' '-O"sl_
   tableList={node2:node2-priv:node2-vip, node3:node3-priv:node3-vip}"' '-O"ret_
   PrivIntrList=<private interconnect list>"' '-O"sl_OHPartitionsAndSpace_
   valueFromDlg={partition and space information}"' '-O-noConfig' '-OPERFORM_
   PARTITION_TASKS=FALSE'
   
   

3. From the %CRS_HOME%\cfgtoollogs directory on the existing node, run the following command:

   <CRS_HOME>\cfgtoollogs\cfgToolAllCommands
   
   

   This instantiates the Virtual Protocol Configuration Assistant (VIPCA), the Oracle RAC Notification Service Utility (racgons), Oracle Clusterware Setup (crssetup), and oifcfg.

7.6.2.2 Cloning Oracle RAC Software on Windows System-Based Environments

Complete the following steps to clone Oracle Database with RAC software on Windows system computers:

1. If you do not have a shared Oracle Database home, then zip the Oracle Database home with Oracle RAC on the existing node and copy it to the new node. Unzip the Oracle Database with Oracle RAC home on the new node in the same directory in which the Oracle Database home with Oracle RAC resided on the existing node. For example, assume that the location of the destination Oracle RAC home on the new node is %ORACLE_HOME%. Skip this step if you have a shared Oracle Database home.

   Note:

   For more information on zipping and unzipping, refer to sections "Source Preparation Phase" and "Cloning Phase".

2. On the new node, go to the %ORACLE_HOME%\clone\bin directory and run the following command where Oracle_Home is the Oracle Database home, Oracle_Home_Name is the name of the Oracle Database home, existing_node is the name of the existing node, and new_node is the name of the new node:

   perl clone.pl ORACLE_HOME=Oracle_Home ORACLE_HOME_NAME=Oracle_Home_Name '-O"CLUSTER_NODES={existing_node,new_node}"' '-OLOCAL_NODE=new_node' '-O-noConfig'
   
   

   If you have a shared Oracle Database home with Oracle RAC, then append the -O-cfs option to the command example in this step and provide a complete path location for the cluster file system. Repeat this step for all nodes.

3. On the new node, run NETCA to add a listener.

4. From the node that you cloned, run DBCA to add the database instance to the new node.

7.6.2.3 Important Considerations when Cloning Oracle Real Application Clusters

Note the following when cloning the Oracle Database with Oracle RAC:

• The order of the nodes specified should always be the same on all hosts.

• Oracle Clusterware should be installed on the cluster nodes prior to starting an Oracle RAC installation.

• For a shared home, you need to also provide a value for the -cfs parameter on the command line.

7.7.1.1 Cloning Oracle Clusterware on UNIX and Linux System-Based Environments

Complete the following steps to clone Oracle Clusterware on UNIX and Linux systems:

1. If you do not have a shared Oracle Clusterware home, then tar the Oracle Clusterware home from an existing node and copy it to the new node. Use $CRS_HOME as the destination Oracle Clusterware home on the new node. If you have a shared Oracle Clusterware home, then skip this step.

   Note:

   For more information on archiving and unarchiving, refer to sections "Source Preparation Phase" and "Cloning Phase".

2. If you do not have a shared Oracle Clusterware home, then on the new node, navigate to the $ORACLE_HOME/clone/bin directory and run the following command where Oracle_home_name is the name of the Oracle home, new_node is the name of the new node, new_node-priv is the private interconnect protocol address of the new node, new_node-vip is the virtual interconnect protocol address of the new node, and central_inventory_location is the location of the Oracle central inventory:

   perl clone.pl ORACLE_HOME=$ORACLE_HOME ORACLE_HOME_NAME=Oracle_home_name '-O"sl_tableList={new_node:new_node-priv:new_node-vip}"' '-O-noConfig''-O"INVENTORY_LOCATION=central_inventory_location"'
   
   

   If you have a shared Oracle Clusterware home, then append the -cfs option to the command example in this step and provide a complete path location for the cluster file system.

   Note:

   Only provide a value for the sl_tableList variable. The Perl clone.pl script takes all other variable settings from the zipped Oracle Clusterware home. This is only true, however, if the source of the zipped home was from an existing node of the cluster that you are extending.

   If you use any other Oracle RAC environment as your cloning source, that is, if you clone from a node in a cluster other than the one that you are extending, then you must provide values for all of the arguments. This includes values for the Oracle Cluster Registry and voting disk location arguments. You must do this because the value for sl_tableList is used as shown in the command example in this step. Also note that you should only specify values for the new node for the sl_tableList options.

3. On the new node, go to the directory that contains the central Oracle inventory. Run the orainstRoot.sh script to populate the file /etc/oraInst.loc with information about the central inventory location.

4. Run the following command on the existing node where new_node is the name of the new node, new_node-priv is the private interconnect protocol address for the new node, and new_node-vip is the virtual interconnect protocol address for the new node:

   $ORACLE_HOME/oui/bin/addNode.sh –silent "CLUSTER_NEW_NODES={new_node}" "CLUSTER_NEW_PRIVATE_NODE_NAMES={new_node-priv}" "CLUSTER_NEW_VIRTUAL_HOSTNAMES={new_node-vip}" –noCopy 
   
   

   Note:

   (Because the clone.pl script has already been run on the new node, this step only updates the inventories on the nodes and instantiates scripts on the local node.

5. On the existing node, run the rootaddnode.sh script from the $ORACLE_HOME/install directory.

6. On the new node, go to the $ORACLE_HOME directory and run the ./root.sh script to start the Oracle Clusterware on the new node.

7. Determine the remote port to use in the next step by running the following command from the $CRS_HOME/opmn/conf directory:

   cat ons.config
   
   

8. From the $CRS_HOME/bin directory on an existing node, run the following command where racgons is the Oracle RAC Notification Service Utility, new_node is the name of the new node, and remote_port is the value from the output of the previous step:

   ./racgons add_config new_node:remote_port
   
   

7.7.1.2 Cloning Oracle RAC Software on UNIX and Linux System-Based Environments

Complete the following steps to clone Oracle Database with RAC software on UNIX and Linux systems:

1. If you do not have a shared Oracle Database home, then tar the Oracle RAC home from the existing node and copy it to the new node. Assume that the location of the destination Oracle RAC home on the new node is $ORACLE_HOME. Otherwise, skip this step.

   Note:

   For more information on archiving and unarchiving, refer to sections "Source Preparation Phase" and "Cloning Phase".

2. If you do not have a shared Oracle Database home, then on the new node go to the $ORACLE_HOME/clone/bin directory and run the following command where existing_node is the name of the node that you are cloning, new_node2 and new_node3 are the names of the new nodes, and Oracle_home_name is the name of the Oracle home:

   perl clone.pl '-OÒCLUSTER_NODES={existing_node,new_node2,new_node3}"' '-O"LOCAL_NODE=new_node2"' ORACLE_HOME=$ORACLE_HOME ORACLE_HOME_NAME=Oracle_home_name '-O-noConfig' 
   
   

   If you have a shared Oracle Database home, then append the -cfs option to the command example in this step and provide a complete path location for the cluster file system.

3. Run the following command on the existing node from the $ORACLE_HOME/oui/bin directory where existing_node is the name of the original node that you are cloning and new_node2 and new_node3 are the names of the new node:

   ./runInstaller –updateNodeList ORACLE_HOME=$ORACLE_HOME ÒCLUSTER_NODES={existing_node,new_node2,new_node3}Ó 
   
   

4. On the new node, go to the $ORACLE_HOME directory and run the following command:

   ./root.sh
   
   

5. On the new node, run Net Configuration Assistant (NETCA) to add a listener.

6. From the node that you cloned, run Database Configuration Assistant (DBCA) to add the new instance.

7.7.2.1 Cloning Oracle Clusterware on Windows System-Based Environments

Complete the following steps to clone Oracle Clusterware on Windows system computers:

1. If you do not have a shared Oracle Clusterware home, then zip the Oracle Clusterware home from the existing node and copy it to the new node. Unzip the home on the new node in the equivalent directory structure as the directory structure in which the Oracle Clusterware home resided on the existing node. For example, assume that the location of the destination Oracle Clusterware home on the new node is %CRS_HOME%. Skip this step if you have a shared Oracle Clusterware home.

   Note:

   For more information on zipping and unzipping, refer to sections "Source Preparation Phase" and "Cloning Phase".

2. On the new node, go to the %CRS_HOME%\clone\bin directory and run the following command where CRS_HOME is the location of the Oracle Clusterware home, CRS_HOME_NAME is the name of the Oracle Clusterware home, and where new_node, new_node-priv, and new_node-vip are the name of the new node, the private interconnect protocol address of the new node, and the virtual interconnect protocol address of the new node respectively:

   perl clone.pl ORACLE_HOME=CRS_HOME ORACLE_HOME_NAME=CRS_HOME_NAME '-O"sl_tableList={new_node:new_node-priv:new_node-vip}"' '-O-noConfig' '-OPERFORM_PARTITION_TASKS=FALSE'
   
   

   If you have a shared Oracle Clusterware home, then append the -O-cfs option to the command example in this step and provide a complete path location for the cluster file system.

   Note:

   Only provide a value for the sl_tableList variable. The Perl clone.pl script takes all other variable settings from the zipped Oracle Clusterware home. This is only true, however, if the source of the zipped home was from an existing node of the cluster that you are extending.

   If you use any other Oracle RAC environment as your cloning source, that is, if you clone from a node in a cluster other than the one that you are extending, then you must provide values for all of the arguments. This includes values for the Oracle Cluster Registry (OCR) and voting disk location arguments. You must do this because the value for sl_tableList is used as shown in the command example in this step. You must also specify the OCR and voting disk locations using the sl_OHPartitionsAndSpace_valueFromDlg variable as well as provide values for the PERFORM_PARTITION_TASKS argument. Only specify values for the new node for the sl_tableList options.

3. Run the following command on the existing node where new_node, new_node-priv, and new_node-vip are the name of the new node, the private interconnect protocol address of the new node, and the virtual interconnect protocol address of the new node respectively:

   %ORACLE_HOME%\oui\bin\addNode.bat -silent "CLUSTER_NEW_NODES={new_node}" "CLUSTER_NEW_PRIVATE_NODE_NAMES={new_node-priv}" "CLUSTER_NEW_VIRTUAL_HOSTNAMES={new_node-vip}" -noCopy -noRemoteActions
   
   

   Note:

   (Because you have already run the clone.pl script on the new node, this step only updates the inventories on the nodes and instantiates scripts on the local node.

4. From the %CRS_HOME%\install directory on the existing node, run the crssetup.add.bat script to instantiate the Virtual Protocol Configuration Assistant (VIPCA) and the Oracle RAC Notification Service Utility (racgons).

7.7.2.2 Cloning Oracle RAC Software on Windows System-Based Environments

Complete the following steps to clone Oracle Database with RAC software on Windows system computers:

1. If you do not have a shared Oracle Database home, then zip the Oracle Database home with Oracle RAC on the existing node and copy it to the new node. Unzip the Oracle Database with Oracle RAC home on the new node in the same directory in which the Oracle Database home with Oracle RAC resided on the existing node. For example, assume that the location of the destination Oracle RAC home on the new node is %ORACLE_HOME%. Skip this step if you have a shared Oracle Database home.

   Note:

   For more information on zipping and unzipping, refer to sections "Source Preparation Phase" and "Cloning Phase".

2. On the new node, go to the %ORACLE_HOME%\clone\bin directory and run the following command where Oracle_Home is the Oracle Database home, Oracle_Home_Name is the name of the Oracle Database home, existing_node is the name of the existing node, and new_node is the name of the new node:

   perl clone.pl ORACLE_HOME=Oracle_Home ORACLE_HOME_NAME=Oracle_Home_Name '-O"CLUSTER_NODES={existing_node,new_node}"' '-OLOCAL_NODE=new_node' '-O-noConfig'
   
   

   If you have a shared Oracle Database home with Oracle RAC, then append the -O-cfs option to the command example in this step and provide a complete path location for the cluster file system.

3. On the existing node, from the RAC_HOME\oui\bin directory run the following command where Oracle_Home is the Oracle Database home with Oracle RAC, existing_node is the name of the existing node, and new_node is the name of the new node:

   setup.exe -updateNodeList ORACLE_HOME=Oracle_Home "CLUSTER_NODES={existing_node,new_node" LOCAL_NODE=existing_node
   
   

4. On the new node, run NETCA to add a listener.

5. From the node that you cloned, run DBCA to add the database instance to the new node.

7.7.3.1 Variables for UNIX and Linux System-Based Environments

Table 7-3 describes the variables that can be passed to clone.pl with -O option for UNIX and Linux systems.

node1    node1-priv    node1-vip
node2    node2-priv    node2-vip

{"eth0:10.87.24.0:2","eth1:140.87.24.0:1","eth3:140.74.30.0:3"}

{"node1:node1-priv:node1-vip:N:Y","node2:node2-priv:node2-vip:N:Y"}.

/oradbshare/oradata/vdisk

/oradbshare/oradata/vdiskmirror1

/oradbshare/oradata/ocr

/oradbshare/oradata/ocrmirror

/oradbshare/oradata/vdiskmirror2

7.7.3.2 Variables for Windows System-Based Environments

Table 7-4 describes the variables that can be passed to clone.pl with -O option for Windows system environments.

CLUSTER_NODES = {"node1"}

{"node1:node1-priv:node1-vip:N:Y","node2:node2-priv:node2-vip:N:Y}

sl_OhPartitionsAndSpace_valueFromDlg = {Disk,Partition,partition size, 0,N/A,1,Disk,Partition, partition size,0,N/A,2,.....)