9.2.2 Using a command to back up and restore definition information (other than PFM - Web Console)
You can back up definition information for Performance Management from the command line without stopping the service. By implementing this process as part of a batch process, you can operate a system in which backups are made automatically.
For Performance Management other than PFM - Web Console, execute the jpccfgbackup command to back up definition information and copy the files to restore the definition information.
- Organization of this subsection
(1) Backing up definition information
Make sure that the following prerequisites are met before you use a command to back up definition information.
- Prerequisites
-
-
To run the script in UNIX, you must be able to use the Korn (ksh).
-
When using a command to back up definition information specific to a certain type of PFM - Agent or PFM - RM, the version of the PFM - Agent or PFM - RM instance must be 10-00 or later. For earlier versions, copy the definition information manually.
-
- Notes:
-
-
When backing up definition information, keep a record of the version numbers of the products in the environment where the backup was taken.
See the Release Notes for details on the product version numbers.
-
Do not specify a backup destination on a network drive.
-
Do not make any changes to definitions or the system configuration while the command is running. The following operations involve changes to definitions or the system configuration:
• When using commands: Any operation that requires administrator role to execute.
• When using PFM - Web Console: Any operation that requires administrator role to execute.
If any such changes are made, the command might fail to execute. If the command provides a return value of 235, execute the command again while leaving the definitions and system configuration unchanged.
-
To back up definition information:
-
Log on to the host where you want to back up definition information.
-
Navigate to the following directory:
installation-folder\tools (in Windows)
/opt/jp1pc/tools (in UNIX)
-
Execute the jpccfgbackup command.
- In a non-cluster system:
-
For example, to back up definition information to the folder C:\backup, execute the command as follows:
jpccfgbackup C:\backup
- In a cluster system:
-
On the executing node:
For example, to back up the definition information for the executing node to the folder C:\backup\jp1-ha1, execute the command as follows with the shared disk connected to the executing node:
jpccfgbackup C:\backup\jp1-ha1
On the standby node:
For example, to back up the definition information for the standby node to the folder C:\backup\jp1-ha2, execute the command as follows without the shared disk connected to the standby node:
jpccfgbackup C:\backup\jp1-ha2
(2) Restoring definition information
Make sure that the following prerequisites are met before you restore definition information backed up by a command.
- Prerequisites
-
-
All the Performance Management services on the host which you want to restore must be stopped.
-
The configuration of the system managed by PFM - Manager must be the same as it was when the backup was taken#.
-
Each host must have the same host name as it did when the backup was taken.
-
The PFM product configuration in the backup environment (the name and version of PFM products, the number of instances of PFM - Agent and PFM - RM, and the name of each instance) must be the same as the PFM product configuration in the environment where the data is being restored.
- #
-
Refers to the following aspects of the instances of PFM - Agent and PFM - RM managed by PFM - Manager:
-
Network configuration (host name, IP address)
-
Instance names, monitoring target names
-
Product versions
-
-
- Note:
-
When you restore Performance Management settings, the version numbers of the products in the restoration environment must exactly match those in the backup environment. See the Release Notes for details on the product version numbers.
To restore definition information:
-
Log on to the host where you want to restore definition information.
-
Make sure that all Performance Management programs and services are stopped on the host.
-
Restore the definition information.
- In a non-cluster system:
-
For example, if the definition files backed up by the command are stored in the C:\backup folder, move the entire contents of C:\backup\jp1pcbackup\localhost to the following directory:
Windows: installation-folder, UNIX: /opt/jp1pc
- In a cluster system:
-
On the executing node:
For example, if the definition files backed up on the executing node by the command are stored in the C:\backup\jp1-ha1 folder, move the entire contents of C:\backup\jp1-ha1\jp1pcbackup\localhost to the following directory:
installation-folder (in Windows)
/opt/jp1pc (in UNIX)
With the shared disk connected to the executing node, move the entire contents of C:\backup\jp1-ha1\jp1pcbackup\logical-host-name#1 to the following directory:
environment-directory#2\jp1pc\ (in Windows)
environment-directory#2/jp1pc/ (in UNIX)
#1
Replace logical-host-name with the name of the logical host where the backup was taken.
#2
environment-directory is the directory created on the shared disk when creating logical hosts.
On the standby node:
For example, if the definition files backed up on the standby node by the command are stored in the C:\backup\jp1-ha2 folder, move the entire contents of C:\backup\jp1-ha2\jp1pcbackup\localhost to the following directory:
installation-folder (in Windows)
/opt/jp1pc (in UNIX)
-
If you have assigned fixed port numbers to programs or services, restore the port number information.
- In a non-cluster system:
-
For example, if the definition files backed up by the command are stored in the C:\backup folder, execute the following command to restore port information:
jpcconf port define -key all -input C:\backup\jp1pcbackup\pfm_port.def
- In a cluster system:
-
After restoring port information on the executing node, do the same for the standby node.
On the executing node:
For example, if the definition files backed up by the command on the executing node are stored in the C:\backup\jp1-ha1 folder, execute the command shown below to restore port information. For logical-host-name, enter the logical host name used by the cluster system.
To restore port information for the physical host:
jpcconf port define -key all -input C:\backup\jp1-ha1\jp1pcbackup\pfm_port.def
To restore port information for the logical host:
jpcconf port define -key all -lhost logical-host-name -input C:\backup\jp1-ha1\jp1pcbackup\pfm_port_logical-host-name.def
After restoring the port information, export the logical host environment definition from the executing node so that the port settings on the standby node can be synchronized with those of the executing node. Execute the command as follows to export port information to the file jp1-ha.conf:
jpcconf ha export -f jp1-ha.conf
On the standby node:
For example, if the definition files backed up by the command on the standby node are stored in the C:\backup\jp1-ha2 folder, execute the following command to restore port information:
To restore port information for the physical host:
jpcconf port define -key all -input C:\backup\jp1-ha2\jp1pcbackup\pfm_port.def
Execute the following command to apply the logical host environment definition of the executing node to the standby node:
jpcconf ha import -f jp1-ha.conf
-
In a non-cluster system, use the jpcspm start command to start the PFM services. In a cluster system, use the cluster software to start the services.
-
If you restored PFM - Manager, execute the following synchronization command from the PFM - Manager to perform synchronization processing for the PFM services of the PFM - Agent or PFM - RM host that connects to the PFM - Manager.
jpctool config sync
For details about the jpctool config sync command, see the chapter that describes commands in the manual JP1/Performance Management Reference.