Important: It is recommended that you select the deployment and configurations that best match your service requirements. All elements must be setup prior to attempting any of the procedures detailed in this chapter. To perform any of the procedures listed in this chapter, you must be logged into the server as a
root user.
•
|
Refer to the System Requirements and Recommendations section in the Overview chapter of this guide to confirm that your system meets the minimum requirements for:
|
Refer to the hard disk partitioning recommendations outlined in the GSS Hardware Sizing and Provisioning Guidelines section in the
Overview chapter of this guide.
Use a text editor to add the following values to the bottom of the system file in the /etc directory and then complete <X-Refs (online)>step <$paranum>
4 before beginning the installation of the GSS application components.
After adding the above values to the system file in the /etc directory, restart the system before installation of the GSS application and components. Enter the UNIX command:
The components that comprise the GSS application software are bundled and distributed in a single compressed file package with a .tar.gz extension. Your sales representative will inform you how to download the appropriate GSS package for your requirements.
Important: Remember that within a procedure, information inside angle brackets
<variable> represents a variable that can be defined by either the user or generated by the system. For example: Create the
/<package> directory to hold the application packages.
gunzip gss_<version>.tar.gz
version is the version number of the GSS software distributed in the compressed tar file. For example, gss_8.0.71.tar.gz.
Step 5
|
Locate the tar file GSS_<version>.tar in the /<packages> directory and untar the file by entering the following command:
|
tar -xvf gss_<version>.tar
During the untar process, a /gss_<version> directory (for example: /gss_8.0.6x) is created in the
/<packages> directory:
Step 6
|
Change to the /<packages>/gss_<version> directory to confirm the presence of the following files:
|
Installation is accomplished using the inst_serv script. It provides a menu-driven interface with question prompts. Most prompts display default values or information derived from the server’s current setup - such as IP addresses for configured interfaces.
•
|
Ctrl-C will abort the installation process at any time during the procedure.
|
Important: It is recommended that you fill in path prompts only after you have created the directories to be used.
Step 1
|
Change to the /<packages>/gss_<version> directory where you stored the GSS application software in step 5 of the previous section.
|
Important: This script will check the version of the operating system installed on the system. If it is not matching the requirements in the
Minimum System Requirements for Stand-alone Deployment section, the script will abort the GSS installation.
Shortly after typing /<install_dir> and pressing
Enter, the following appears:
Important: All values that appear initially for this menu are system defaults and you do not need to make changes if the values are acceptable.
Important: The GSS and FileGen are set to run in archive mode by default. The archive mode, used in deployments that do not include CGFs, instructs the server to save records to a file.
Important: Note that the script has sensed both the number of interfaces and their IP addresses.
Pressing n completes the menu-driven portion of the installation process and displays the configuration that you have created.
Step 8
|
Press y to return to the menus and change the configuration or press n to continue the installation process.
|
0 * * * * /<install_dir>/gss/bin/cleanup.sh >> /<
install_dir>/gss/log/cleanup.log 2>&1
Step 9
|
Enter n to complete the installation without starting the GSS. Press return to accept the yes default to complete the installation process and start the GSS.
|
The status display indicates that GSS, FileGen, PSMON, and PostgreSQL have all been started. If nothing displays, turn to the Troubleshooting the GSS section in the
GTPP Storage Server Administration chapter. In most cases, if the other components are started, then the PostgreSQL has also been started.
Before working with the GSS, it is recommended to create a write-protected copy of the gss.cfg file and store it in a separate directory. To ensure you remember the configuration for your software version, we suggest that you store the in the
/<packages>/<gss_version directory.
Installation is accomplished using the inst_serv script. It provides a menu-driven interface with question prompts. Most prompts display default values or information derived from the server’s current setup - such as IP addresses for configured interfaces.
•
|
Ctrl-C will abort the installation process at any time during the procedure.
|
Important: It is recommended that you fill in path prompts only after you have created the directories to be used.
Step 1
|
Change to the /<packages>/gss_<version> directory where you stored the GSS application software.
|
Important: This script will check the version of operating system and cluster software installed on the system. If it is not matching the requirements in the
Minimum System Requirements for Cluster Deployment section, the script will abort the GSS installation.
Important: Note that the script senses whether the server is a Stand-alone node or a Cluster node. If you made changes to the system file as in the
Installation First Steps section, then you can continue. If not, abort the installation using CTRL+C, make the changes to the system file, and then reboot. After rebooting, begin the installation procedure again.
Step 3
|
Enter y(yes) to continue the installation.
|
Step 4
|
Press Enter to accept the default directory /sharedgss/gss or enter the name of another directory. Next you are prompted for the location to install the GSS.
|
Important: In case of cluster mode, it is highly recommended that you do not install the GSS application in the
/opt,
/opt/gss/, or
/opt/postgres directory.
Step 5
|
Press Enter to accept the default or enter the name of the directory where the GSS active components are to be installed. It is recommended that you put this directory at the root level. The installation script creates the directory if needed.
|
Important: This enables backup mode for the GSS node in a cluster deployment.
Step 6
|
Enter y(yes) to enable node switchover. The installation continues with a menu to configure the PostgreSQL parameters for backup.
|
Step 9
|
Enter n to save changes or defaults and move to the next menu.
|
Important: The GSS and FileGen are set to run in archive mode by default. The archive mode, used in deployments that do not include CGFs, instructs the server to save records to a file.
Important: The
Specify the GTPP Dictionary option appears only if the
Support for LRSN rewrite or
Encoding of IP Address in binary format parameter is enabled. Otherwise, the GTPP dictionary is set to default.
Pressing n completes the menu-driven portion of the installation process and displays the configuration that you have created.
Step 15
|
Press y to return to the menus and change the configuration or press n or Enter to continue the installation process.
|
Step 16
|
Enter y(yes) to continue the installation.
|
Step 1
|
Change to the /<packages>/gss_<version> directory where you stored the GSS application software.
|
Important: This script will check the version of operating system and cluster software installed on the system. If it is not matching the requirements in the
Minimum System Requirements for Cluster Deployment section, the script will abort the GSS installation.
Step 3
|
Enter y(yes) to perform GSS installation in cluster mode and continue the installation.
|
Shortly after typing /<shared_gss_dir> and pressing
Enter, the configuration created for cluster node 1 appears:
Step 5
|
Press Enter to continue the installation.
|
Step 6
|
Enter y (yes) to continue the installation.
|
Step 7
|
Enter y (yes) to continue the installation.
|
The status display indicates that GSS, FileGen, PSMON, and Postgres have all been started. If nothing displays, turn to the Troubleshooting the GSS section in the
GTPP Storage Server Administration chapter. In most cases, if the other components started, then the PostgreSQL has also been started.
Before doing anything with the GSS, it is recommended to create a write-protected copy of the gss.cfg file and store it in a separate directory. To ensure you remember the configuration for your software version, we suggest that you store it in the
/<packages>/<gss_version> directory.
Step 4
|
Enter y(yes) to continue the uninstall.
|
Step 4
|
Enter y (yes) to continue the uninstall and start the cleanup operation.
|
Step 5
|
Enter y (yes) to continue.
|
Step 6
|
Enter y (yes) to continue.
|
•
|
Change to /<clus_install_dir>/gss/etc and make a copy of the current gss.cfg.
|
gunzip gss_<version>.tar.gz
<version> is the version number of the GSS software distributed in the compressed tar file. For example, gss_8.0.xx.tar.gz.
Step 4
|
Locate the tar file GSS_<version>.tar in the /<packages> directory and untar the file by entering the following command:
|
tar -xvf gss_<version>.tar
During the untar process, a /gss_<version> directory (for example: /gss_8.0.xx) is created in the
/<packages> directory.
Installation is accomplished using the inst_serv script. It provides a menu-driven interface with question prompts. Most prompts display default values or information derived from the server’s current setup - such as IP addresses for configured interfaces.
•
|
Ctrl-C will abort the installation process at any time during the procedure.
|
Important: It is recommended that you fill in path prompts only after you have created the directories to be used.
Step 1
|
Change to the /<packages>/gss_<version> directory that was created when the files were uncompressed.
|
Important: This script will check the version of the operating system installed on the system. If it is not matching the requirements in the
Minimum System Requirements for Stand-alone Deployment section, the script will abort the GSS installation.
Important: If the upgrade is to a GSS 8.0.xx then
/opt will appear as a default value for the
GSS Installation dir parameter.
Step 5
|
Enter y (yes) to continue the upgrade.
|
Step 6
|
Enter y or press Enter to start the GSS and all related processes (e.g. the Process Monitor).
|
•
|
Change to /<clus_install_dir>/gss/etc directory and make a copy of the current gss.cfg.
|
gunzip gss_<version>.tar.gz
<version> is the version number of the GSS software distributed in the compressed tar file. For example, gss_8.0.xx.tar.gz.
Step 4
|
Locate the tar file GSS_<version>.tar in the /<packages> directory and untar the file by entering the following command:
|
tar -xvf gss_<version>.tar
During the untar process, a /gss_<version> directory (for example: /gss_8.0.xx) is created in the
/<packages> directory.
Important: This script will check the version of operating system and cluster software installed on the system. If it is not matching the requirements in the
Minimum System Requirements for Cluster Deployment section, the script will abort the GSS installation.
Important: The database can only be upgraded on the node where the GSS is currently online (primary).
Step 5
|
Press Enter or type y to reboot into stand-alone mode and continue the upgrade process.
|
Important: It will take a minute or two for the resource group to switch to node2 and for node1 to reboot in stand-alone mode.
Step 9
|
Press Enter or type y (yes) to continue the upgrade.
|
Step 10
|
Press Enter or type y to confirm the upgrade of the cluster resources.
|
Important: ./serv version is used when upgrading to GSS versions higher than x.x.69.
Important: Reminder from previous step -
/opt is a default installation directory for some versions of GSS.
Important: ./serv switch is used when upgrading to GSS versions higher than x.x.69.
Step 16
|
Perform step 1 to step 11 on node2 to upgrade it to the newer version of GSS.
|
Important: While executing
step 3 on the second node, do not choose to upgrade database, as it has already upgraded for first node. After completing the steps on node2, GSS upgrade on cluster setup will be completed.
The GSS installer installs all the binaries for GSS, PostgreSQL and StarentGSS (cluster resource type) package in /opt/gss_global directory. The installer has to be executed on both the nodes of the cluster. The
inst_serv script is used to create instances after initial installation. Each instance will have a separate configuration file, log directory, and tools directory. During uninstallation, each instance can be uninstalled separately and will not have any impact on the other instances. Global installation can be only uninstalled if there are no instances configured or running on the system.
Installation is accomplished using the inst_serv script to install GSS binaries. It provides a menu-driven interface with question prompts. Most prompts display default values or information derived from the server’s current setup - such as IP addresses for configured interfaces.
•
|
Ctrl-C will abort the installation process at any time during the procedure.
|
Important: It is recommended that you fill in path prompts only after you have created the directories to be used.
Step 1
|
Change to the /<packages>/gss_<version> directory where you stored the GSS application software in step 5 of the previous section.
|
Important: This script will check the version of the operating system installed on the system. If it is not matching the requirements in the
Minimum System Requirements for Stand-alone Deployment section, the script will abort the GSS installation.
If you made changes to the system file as in the Installation First <X-Refs (online)>step <$paranum>s section, then continue to the next step.
Shortly after typing /<install_dir> and pressing
Enter, the following appears:
Important: All values that appear initially for this menu are system defaults and you do not need to make changes if the values are acceptable.
Important: The GSS and FileGen are set to run in archive mode by default. The archive mode, used in deployments that do not include CGFs, instructs the server to save records to a file.
Important: Note that the script has sensed both the number of interfaces and their IP addresses.
Pressing n completes the menu-driven portion of the installation process and displays the configuration that you have created.
Step 8
|
Press y to return to the menus and change the configuration or press n to continue the installation process.
|
0 * * * * /<install_dir>/gss/bin/cleanup.sh >> /<
install_dir>/gss/log/cleanup.log 2>&1
Step 9
|
Enter n to complete the installation without starting the GSS. Press return to accept the yes default to complete the installation process and start the GSS.
|
The status display indicates that GSS, FileGen, PSMON, and PostgreSQL have all been started. If nothing displays, turn to the Troubleshooting the GSS section in the
GTPP Storage Server Administration chapter. In most cases, if the other components are started, then the PostgreSQL has also been started.
Before working with the GSS, it is recommended to create a write-protected copy of the gss.cfg file and store it in a separate directory. To ensure you remember the configuration for your software version, we suggest that you store the in the
/<packages>/<gss_version directory.
Installation is accomplished using the inst_serv script. It provides a menu-driven interface with question prompts. Most prompts display default values or information derived from the server’s current setup.
•
|
Ctrl-C will abort the installation process at any time during the procedure.
|
The information from the prompts is used to generate the GSS configuration file (gss.cfg). Using a text editor, this file can be changed at anytime after the installation.
Step 1
|
Change to the /<packages>/gss_<version> directory where you stored the GSS application software.
|
Important: This script will check the version of operating system and cluster software installed on the system. If it is not matching the requirements in the
Minimum System Requirements for Cluster Deployment section, the script will abort the GSS installation.
Important: Note that the script senses whether the server is a Stand-alone node or a Cluster node. If you made changes to the system file as in the
Installation First Steps section, then you can continue. If not, abort the installation using
CTRL+C, make the changes to the system file, and then reboot. After rebooting, begin the installation procedure again.
Step 3
|
Enter y (yes) to continue the installation.
|
Step 4
|
Press Enter to accept the default directory /sharedgss/gss or enter the name of another directory. Next you are prompted for the location to install the GSS.
|
Important: In case of cluster mode, it is highly recommended that you do not install the GSS application in the /opt, /opt/gss/, or /opt/postgres directory.
Step 5
|
Press Enter to accept the default or enter the name of the directory where the GSS active components are to be installed. It is recommended that you put this directory at the root level. The installation script creates the directory if needed.
|
Important: This enables backup mode for the GSS node in a cluster deployment.
Step 6
|
Enter y (yes) to enable node switchover. The installation continues with a menu to configure the PostgreSQL parameters for backup.
|
Step 9
|
Enter n to save changes or defaults and move to the next menu.
|
Important: The GSS and FileGen are set to run in archive mode by default. The archive mode, used in deployments that do not include CGFs, instructs the server to save records to a file.
Important: The
Specify the GTPP Dictionary option appears only if the
Support for LRSN rewrite or
Encoding of IP Address in binary format parameter is enabled. Otherwise, the GTPP dictionary is set to default.
Pressing n completes the menu-driven portion of the installation process and displays the configuration that you have created.
Step 15
|
Press y to return to the menus and change the configuration or press n or Enter to continue the installation process.
|
Step 16
|
Enter y (yes) to continue the installation.
|
Step 1
|
Change to the /gss_<version> directory where the GSS files are stored.
|
Important: This script will check the version of operating system and cluster software installed on the system. If it is not matching the requirements in the
Minimum System Requirements for Cluster Deployment section, the script will abort the GSS installation.
Step 3
|
Run /opt/gss_clobal/make_gss_instance to create a GSS instance.
|
Step 4
|
Enter y (yes) to continue the installation.
|
Shortly after typing /<shared_gss_dir> and pressing
Enter, the configuration created for cluster node 1 appears.
Step 6
|
Press Enter to continue the installation.
|
Step 7
|
Enter y to create another GSS instance. Repeat the above steps in order to add and modify the new GSS instance with configurations as required.
|
Step 8
|
Enter n if you want to create more instances.
|
The StandaloneGSSUninstall.sh uninstallation script is the script used for uninstall process on a standalone mode. This script created in each GSS instance installation directory. For example, GSS instance installation directory is
/cvserver then, uninstallation script is created in
/cvserver as
StandaloneGSSUninstall.sh.
Step 3
|
Enter y (yes) to continue the uninstall process.
|
The system goes to the installation directory where the system-generated uninstall_config_file resides and uses the information in that file to complete the install process.
Step 5
|
Enter y (yes) to continue the uninstall process.
|
Step 7
|
Enter y (yes) to continue the uninstall process.
|
Step 8
|
Enter y (yes) to remove the installed package instance.
|
Step 9
|
Enter y (yes) to continue the uninstall process.
|
The ClusterGSSUninstall.sh uninstallation script is the script used for uninstall process on a standalone mode. This script created in each GSS instance installation directory. For example, GSS instance installation directory is
/cvserver then, uninstallation script is created in
/cvserver as
ClusterGSSUninstall.sh. During both the installation and upgrade processes, the system generates system-specific uninstall files: an uninstall script and an uninstall configuration file.
Step 3
|
Enter y (yes) to continue the uninstall and start the cleanup operation.
|
Step 5
|
Enter y (yes) to continue the uninstall process.
|
Step 7
|
Enter y (yes) to continue the uninstall and start the cleanup operation.
|
Step 8
|
Enter y (yes) to continue.
|
Step 9
|
Enter y (yes) to continue.
|
If the following message is displayed when Global_Gss_Unistall.sh is run, then some instance which needs to be uninstalled is still present and then, Global instance needs to be uninstalled. To identify which instance needs to be uninstalled, please see
/etc/gss/.instance_configfile.
•
|
Change to /<clus_install_dir>/gss/etc directory and make a copy of the current gss.cfg.
|
gunzip gss_<version>.tar.gz
<version> is the version number of the GSS software distributed in the compressed tar file. For example, gss_8.0.xx.tar.gz.
Step 4
|
Locate the tar file GSS_<version>.tar in the /<packages> directory and untar the file by entering the following command:
|
tar -xvf gss_<version>.tar
During the untar process, a /gss_<version> directory (for example: /gss_8.0.xx) is created in the
/<packages> directory.
Installation is accomplished using the inst_serv script. It provides a menu-driven interface with question prompts. Most prompts display default values or information derived from the server’s current setup - such as IP addresses for configured interfaces.
•
|
Ctrl-C will abort the installation process at any time during the procedure.
|
Important: It is recommended that you fill in path prompts only after you have created the directories to be used.
Step 1
|
Change to the /<packages>/gss_<version> directory that was created when the files were uncompressed.
|
Important: This script will check the version of the operating system installed on the system. If it is not matching the requirements in the
Minimum System Requirements for Stand-alone Deployment section, the script will abort the GSS installation.
Important: If the upgrade is to a GSS 8.0.xx then
/opt will appear as a default value for the
GSS Installation dir parameter.
Step 5
|
Enter y (yes) to continue the upgrade.
|
Step 6
|
Enter y or press Enter to start the GSS and all related processes (e.g. the Process Monitor).
|
•
|
Change to /<clus_install_dir>/gss/etc directory and make a copy of the current gss.cfg.
|
gunzip gss_<version>.tar.gz
<version> is the version number of the GSS software distributed in the compressed tar file. For example, gss_8.0.xx.tar.gz.
Step 4
|
Locate the tar file GSS_<version>.tar in the /<packages> directory and untar the file by entering the following command:
|
tar -xvf gss_<version>.tar
During the untar process, a /gss_<version> directory (for example: /gss_8.0.xx) is created in the
/<packages> directory.
Important: This script will check the version of operating system and cluster software installed on the system. If it is not matching the requirements in the
Minimum System Requirements for Cluster Deployment section, the script will abort the GSS installation.
Important: The database can only be upgraded on the node where the GSS is currently online (primary).
Step 5
|
Press Enter or type y to reboot into stand-alone mode and continue the upgrade process.
|
Important: It will take a minute or two for the resource group to switch to node2 and for node1 to reboot in stand-alone mode.
Step 9
|
Press Enter or type y (yes) to continue the upgrade.
|
Step 10
|
Press Enter or type y to confirm the upgrade of the cluster resources.
|
Important: ./serv version is used when upgrading to GSS versions higher than x.x.69.
Important: Reminder from previous step -
/opt is a default installation directory for some versions of GSS.
Important: ./serv switch is used when upgrading to GSS versions higher than x.x.69.
Step 16
|
Perform step 1 to step 11 on node2 to upgrade it to the newer version of GSS.
|
Important: While executing
step 3 on the second node, do not choose to upgrade database, as it has already upgraded for first node. After completing the steps on node2, GSS upgrade on cluster setup will be completed.
Important: IPMP is a feature supported on Sun® Solaris® provided by Sun Microsystems. The configuration is included in the
System Administration Guide. For more information, refer to the
Sun documentation.
•
|
Test Addresses: These are IP addresses assigned to each board/interface of the multipath group, they do not move but should not be used for connections in or out of the host.
|
•
|
Multipath/float Address: This is the IP address allocated to a Multipath Interface Group that is shared between all devices in the group (either by load sharing or active standby).
|
•
|
<NIC_1> and <NIC_2> are the network interface devices; i.e. bge0, bge1 etc.
|
Step 2
|
Create an <NIC_1> for the Standby network device with the following entry:
|
<hostname>-
<NIC_1> netmask <netmask> broadcast+group <multipath_grp> deprecated -failover standby up
<hostname> is name of the Host and
<NIC_1> is the network device to be set as Standby.
<multipath_grp> is Multipath Interface Group name given to the group of network devices in a multipath configuration.
<netmask> is the sub-netmask used by network.
Step 3
|
Create an /etc/hostname.<NIC_2> for the active network device with the following entry:
|
<hostname>-
<NIC_2> netmask 255.255.255.0 broadcast+group <multipath_grp>deprecated -failover up addif <hostname>-
active netmask 255.255.255.0 broadcast+failover up
<hostname> is name of the Host and
<NIC_2> is the network device to be set as active.
<multipath_grp> is Multipath Interface Group name given to the group of network devices in a multipath configuration.
Step 4
|
Edit the /etc/hosts file using “vi editor” and add the following three entries:
|
<multipath_IP_address> <
hostname>-active
<test_IP_address_NIC_1> <
hostname>-
<NIC_1>
<test_IP_address_NIC_2> <
hostname>-
<NIC_2>
multipath_IP_address is the IP address allocated to a Multipath Interface Group that is shared between all devices in the group (either by load sharing or Active-Standby).
test_IP_address_NIC_1 is the IP addresses assigned to
<NIC_1> interface of the multipath group, they do not move but should not be used for connections in or out of the host.
test_IP_address_NIC_2 is the IP addresses assigned to
<NIC_2> interface of the multipath group, they do not move but should not be used for connections in or out of the host.
•
|
<NIC_1> and <NIC_2> are the network interface devices; i.e. bge0, bge1 etc.
|
•
|
<my_address> is associated with Multipath IP address < multipath_IP_address> in the /etc/hosts file
|
Step 2
|
Create an /etc/hostname.<NIC_1> for the Active network device with the following entry:
|
<my_address> netmask + broadcast + group <multipath_grp> up
<my_address> is associated with Multipath IP address
<multipath_IP_address> in the
/etc/host file.
<multipath_grp> is Multipath Interface Group name given to the group of network devices in a multipath configuration.
Step 3
|
Create an /etc/hostname.<NIC_2> for the Standby network device with the following entry:
|