3.3.4 Notes on installing and uninstalling JP1/Base
- Organization of this subsection
(1) Installation
-
If you see a message stating that installation failed while using the Hitachi Program Product Installer, the installation log is output to the /etc/.hitachi/.hitachi.log file. We recommend that you back up this file as required because this file is overwritten every time you start the Hitachi Program Product Installer.
The installation log is output to the /tmp/HITACHI_JP1_INST_LOG directory.
-
If you are installing JP1/Base in a Solaris non-global zone, use a version that supports the non-global zone (09-00 or later) for all JP1/Base instances on the same device.
-
Only install this product on a local disk. Do not install this product on network connection disks (NFS, NAS, and others).
(2) Re-installation
-
If you are reinstalling JP1/Base over the existing JP1/Base program, be sure to quit JP1/Base and all JP1 programs, and quit all programs currently accessing the JP1/Base event service.
If you are using JP1/AJS - Manager, stop the JP1/AJS - Monitor service.
-
When you overwrite an existing JP1/Base program with a newer version, the Hitachi Network Objectplaza Trace Library (HNTRLib2) is disabled. You cannot collect information with the integrated trace log even when you run JP1/Base. When JP1/Base is overwritten, you should use the ps command to check that the Hitachi Network Objectplaza Trace Library (HNTRLib2) is activated (it is activated if the hntr2mon process is running). If not, use the hntr2mon command to run it. For details on the hntr2mon command, see hntr2mon (UNIX only) in 15. Commands.
-
When uninstalling JP1/Base and then reinstalling it, you must first uninstall JP1/Base and all products that require it. Then, reinstall JP1/Base and then the products that require it.
-
JP1/IM - Manager
Reinstall JP1/Base and then re-set up JP1/Base and JP1/IM - Manager.
-
JP1/AJS
Reinstall JP1/Base and then re-set up JP1/Base and JP1/AJS.
-
JP1/AJS2 for Mainframe
Uninstall both JP1/Base and JP1/AJS2 for Mainframe and then reinstall JP1/Base and JP1/AJS2 for Mainframe. Next, re-set up JP1/Base and JP1/AJS2 for Mainframe.
-
JP1/Power Monitor
Reinstall JP1/Base and then re-set up JP1/Base and JP1/Power Monitor. However, you do not need to set up JP1/Power Monitor again, if you have not set up a logical host or linkage with JP1/AJS.
-
-
If reinstalling JP1/Base after forcibly uninstalling it without uninstalling the product that requires JP1/Base, it is necessary to restore the common definition information other than JP1/Base that has been backed up. For the procedure to restore backed-up common definition information, see (8) Procedure for restoring common definition information of a JP1 product that requires JP1/Base.
-
If you are using the SNMP trap converter, execute the imevtgw_setup command after reinstalling JP1/Base.
(3) Setting the OS environment
At JP1/Base installation, the port numbers listed in C. List of Port Numbers are added to the /etc/services file. This information is removed when JP1/Base is uninstalled.
(4) Uninstallation
-
After uninstalling JP1/Base, check whether the following directories still exist and delete them if so:
-
/etc/opt/jp1base
-
/opt/jp1base
-
/var/opt/jp1base
-
-
If you uninstall only JP1/AJS after JP1/AJS and JP1/Base are installed, the event service might not start up. In this case, you should remove the include ajs-conf parameter lines or change them to comments (add # to the beginning of the lines) in the event server settings file (conf).
-
The following installer log file is created. Delete this log file after the installation ends normally.
/tmp/HITACHI_JP1_INST_LOG/jp1base_inst{1|2|3|4|5}.log
-
The port numbers for the jesrd service are not removed from the services file. Delete them if they are no longer needed.
-
If any of the following products is installed, the KAVA1870-E message is output, and uninstallation cannot be performed.
-
JP1/AJS3 - Manager
-
JP1/AJS3 - Agent
-
JP1/AJS3 - Agent Minimal Edition
-
JP1/AJS3 - WOA
-
JP1/Power Monitor
-
JP1/IM2 - Manager
-
JP1/IM2 - View
-
JP1/IM2 - EG
-
(5) Overwrite installation
Note the following if you are performing an overwrite installation of JP1/Base in an environment running an earlier version of a JP1 program:
-
You cannot install JP1/Base Version 7 or later on a host running Version 6 of JP1/IM - Central Console or JP1/AJS.
-
The storage format of the command execution log (ISAM) files has changed in Version 8. If you are using JP1/IM and you upgraded to JP1/Base Version 8 or later by overwriting JP1/Base 07-51 or earlier, make sure that you execute the jcocmdconv command before you recommence JP1/IM operation.
The jcocmdconv command migrates the command execution log (ISAM) files accumulated in a previous version of JP1/Base to the file format used in Version 8 or later. If you do not execute this command, you will not be able to access the command execution logs accumulated in Version 7 or earlier. During cluster operation, while the shared disk can be accessed, execute the jcocmdconv command once only (specifying the logical host) on either the primary or secondary node.
For details on the jcocmdconv command, see jcocmdconv in 15. Commands.
A command execution log is created only in JP1/Base on the manager host (on which JP1/IM is also installed).
-
In Version 9, the save-rep flag has been added to the options parameter in the event server settings file (conf). Setting this flag saves the duplication prevention table of the event database into the file. If this flag is not set, the duplication prevention table is saved to memory. In this case, if the event server is restarted, the table is deleted, and then re-created, causing the database to take longer to receive JP1 events forwarded from other hosts. We recommend that you set the save-rep flag for the event server that receives JP1 events forwarded from other hosts.
If you perform an overwrite installation from JP1/Base 08-00 and earlier, this flag will not be set. In this case, you must perform the following procedure to create the duplication prevention table in the file.
To create this table in the file:
-
Add the save-rep flag to the options parameter in the event server settings file.
For details on the event server settings file, see Event server settings file in 16. Definition Files.
-
Execute the jevdbmkrep command.
For details on the jevdbmkrep command, see jevdbmkrep in 15. Commands.
-
Start the event server.
-
-
The log-file trap startup definition file (jevlog_start.conf) is new in version 10-00. If you link with version 10-00 or later of JP1/IM, you can use the IM configuration management feature in conjunction with this file to control the starting and stopping of log file traps.
If you intend to use IM configuration management to control the starting and stopping of log file traps, and have defined the startup of a log file trap in jbs_start, use the following procedure to migrate the definition to the log-file trap startup definition file.
-
Delete the startup definition for the log file trap from jbs_start.
-
In the log-file trap startup definition file, enter the settings for the log file trap that you want to stop and start.
For details on the log-file trap startup definition file, see Log-file trap startup definition file in 16. Definition Files.
Note the following when migrating these definitions:
-
You must specify the name of the log-file trap startup definition file in the -f option of the jevlogstart command.
-
Do not specify a monitoring target name or log-file trap action definition file name that is already being used by another log file trap. If the name is already in use, choose another.
-
-
In versions 11-50 or later, the number of records in the command execution log file (initial setting is 20,000 records) may be insufficient in case of a large-scale system (1,024 or more agents) to use JP1/IM - Manager's remote commands or automated actions, and therefore we recommend that you expand the setting of the number of records.
If the function for automatically re-organizing the command execution log file is enabled however, the amount of time required for startup might increase in proportion to the file size. Therefore, confirm beforehand that the newly specified number of records does not hinder operation. For details see 4.9 Function for automatically re-organizing the command execution log file.
-
If you are reinstalling JP1/Base over the existing JP1/Base program, be sure to quit JP1/Base and all JP1 programs, and quit all programs currently accessing the JP1/Base event service.
If you are using JP1/AJS - Manager, stop the JP1/AJS - Monitor service.
-
If there is only one FQDN format event server name in the Event server index file for the JP1/Base physical host and you call the JP1 event acquisition function, add the definition of * (indicating that the event server name runs on local host) as event-server-name. (This works on JP1/Base version 12-10 or later.) If you do not add the definition mentioned above, the JP1 event acquisition function fails.
For details, see Event server settings file of 16. Definition Files.
(6) Notes on performing installation using disk replication (Linux only)
This is a Hitachi program product that supports the installation of replicated disks by JP1/ServerConductor/Deployment Manager and Hitachi Compute Systems Manager Deployment Manager Plug-in, or replication functionality that uses image files provided by virtualized platforms.
When performing disk replication installation, we recommend that you replicate the default settings of each JP1/Base definition file after installation and setup.
However, if you change the settings and then use the same settings in the target replication hosts, note the following points.
(a) Precautions for before replication
- Setting the processing environment of the event service
-
Do not set IP addresses or host names for each parameter in the event server index file, the event server settings file, or the API settings file.
- Deleting definition information for system configurations
-
When defining the target replication host under the system configuration in JP1/IM, delete the previously set configuration information by using the jbsrt_del command.
(b) Precautions for after replication
- Initializing the event database
-
JP1 events might remain in the event database if you restart the computer with an internally generated host name at the time of disk replication installation. Initialize the event database before operation starts.
For operational procedures, see the JP1/ServerConductor/Deployment Manager User's Guide and Hitachi Compute Systems Manager Deployment Manager Plug-in User's Guide. For details about the replication functionality of virtualized platforms, see the product manual.
(7) Notes on the silent installation functionality
This software product can be installed by using the silent installation functionality.
- Execute the following command:
-
In HP-UX:
Supplied-media/IPFHPUX/setup -f -k Product-type-No Path-of-supplied-media
In Solaris:
Supplied-media/SOLARIS/setup -f -k Product-type-No Path-of-supplied-media
In AIX:
Supplied-media/AIX/setup -f -k Product-type-No Path-of-supplied-media
In Linux:
Supplied-media/LINUX/setup -f -k Product-type-No Path-of-supplied-media
- Notes on performing silent installation
-
-
To execute the contents of supplied media directly from a hard drive, copy the contents of the supplied media to a directory on the hard drive whose path does not include any spaces. Then, compare the copied files to the original files (at the binary level, for example) to ensure that they are the same.
-
Check the return value of the execution result to verify that the installation ended successfully. For details about the return values, see the JP1 website.
-
(8) Procedure for restoring common definition information of a JP1 product that requires JP1/Base
The following describes the procedure for restoring common definition information when reinstalling JP1/Base without uninstalling the JP1 product that requires JP1/Base.
Note: Do not start the JP1/Base service or other JP1 product service during the restore procedure.
And execute the following command to backup the common definition information of all JP1 products in case of performing incorrect procedure to restore the common definition information.
-
Physical host:
jbsgetcnf > backup-file-name
-
Logical host:
jbsgetcnf -h logical-host-name > backup-file-name
-
Back up JP1/Base.
For details on backups, see 3.5.3 Backup and recovery (in UNIX).
If JP1/IM - Manager is installed, backup the command execution log (ISAM) file. For details about the backup of command execution log (ISAM) file, see the manual JP1/Integrated Management 2 - Manager Administration Guide (Database backup and recovery).
Collect the log data before you perform the procedure to restore the common definition information in case the integrated trace log is deleted. For details on collecting log data, see 18.4 How to collect data.
-
Check the output destination of the common definition information backup file.
Check whether there is sufficient free disk space (1 + 0.4 x number-of-logical-hosts [MB] or more) for the output destination of the common definition information backup file (\tmp\). And check whether the common definition information backup file and output directory have not been locked by applications like editor. And also check for the access permission.
If there is insufficient free disk space, the common definition information backup file will not be output correctly. For example, the file size of the common definition information backup becomes 0 byte.
If directory or file has been locked by some application or if you do not have access permission, error message KAVA1871-E will be output during the uninstallation.
-
Set the following environment variables:
-
# JP1_BASE_UNINST_FORCE=1
-
# export JP1_BASE_UNINST_FORCE
-
-
Uninstall JP1/Base.
Uninstall JP1/Base as prompted by the Hitachi Program Product Installer by referring 3.3.3 Uninstalling JP1/Base.
Uninstallation can be performed even if the prerequisite products of JP1/Base are installed.
During the uninstallation, common definition information excluding JP1/Base is output as the common definition information backup file.
-
Physical host: /tmp/JP1_COMDEF/comdef_JP1_DEFAULT.conf
-
Logical host: /tmp/JP1_COMDEF/comdef_logical-host-name.conf
-
-
Clear the setting of the JP1_BASE_UNINST_FORCE environment variable.
-
Reinstall JP1/Base.
For details on install, see 3.3.1 Installing JP1/Base.
-
Re-set up JP1/Base.
If you are operating in a cluster, create a logical host.
The logical host name (Uppercase/Lowercase) and shared directory name cannot be changed.
Recover the adapter command configuration file (/opt/jp1base/plugin/conf/*.conf) from the backup in step 1.
If JP1/IM - Manager is installed, recover the command execution log (ISAM) file from the backup in step 1.
Recover the event database from the backup in step 1. For details on recovering the event database, see 3.5.3(4)(a) When the forwarding settings file is not configured to forward JP1 events to other hosts.
Build the JP1/Base environment.
-
When recovering the JP1/Base environment:
Recover the JP1/Base other definition files, common definition information, and jp1hosts2 information from the backup in step 1.
For details on recovery, see 3.5.3 (3) Recovering JP1/Base setup information.
-
When rebuilding JP1/Base environment:
Perform JP1/Base setup again.
For details on setup, see 3.4 JP1/Base setup.
And also, perform other settings as necessary.
If you have set up JP1/Base in the setup of another product, set it up again.
-
-
Execute the jbssetcnf command using the common definition information backup file output in step 4.
Restore the common definitions for JP1 products excluding JP1/Base. Execute the command for the physical host and all logical hosts.
jbssetcnf common-definition-information-backup-file-name
-
Delete the common definition information backup file output directory in step 4.
Delete the common definition information backup file output directory (\tmp\JP1_COMDEF\).