Job Management Partner 1/Automatic Job Management System 3 Configuration Guide 1

[Contents][Glossary][Index][Back][Next]


16.2.2 Procedure for setting up the environment for a cluster system

Use commands to define the items described in 16.2.1 Environment setup items for a cluster system. Do this on both the primary and secondary nodes.

This subsection describes the tasks related to the environment setting commands that must be performed on both the primary and secondary nodes. Stop all the JP1 services and then set up the environment in JP1/AJS3 - Manager or JP1/AJS3 - Agent.

The following figure provides an overview of setting up the environment.

Figure 16-2 Overview of setting up the environment

[Figure]

Organization of this subsection
(1) Tasks required on the primary node (JP1/AJS3 - Manager)
(2) Tasks required on the secondary node (JP1/AJS3 - Manager)
(3) Tasks required on the primary node (JP1/AJS3 - Agent)
(4) Tasks required on the secondary node (JP1/AJS3 - Agent)

(1) Tasks required on the primary node (JP1/AJS3 - Manager)

To set up a cluster system environment for JP1/AJS3 - Manager on the primary node:

  1. Perform the JP1/Base tasks required on the primary node.
    For details about the JP1/Base tasks, see the Job Management Partner 1/Base User's Guide.
    Note that, depending on the network configuration or operating environment, you might need to set the jp1hosts information. Specify the necessary settings by referring to the notes on cluster system operation and the procedure for setting the jp1hosts information in the Job Management Partner 1/Base User's Guide.
  2. Set up the physical host.
    For details about how to set up the physical host, see 12.1.2 Setting up JP1/AJS3 - Manager.
    To use JP1/AJS3 Console Manager and JP1/AJS3 Console Agent, execute the following commands:
     
    /opt/jp1ajs2cm/bin/jp1ajs2cmsetup
    /opt/jp1ajs2/bin/jp1ajs2casetup
     
    You do not have to execute the above commands if the physical host has been set up. For the JP1/AJS3 Console service, automatic startup and termination must be set. For details about how to set automatic startup and termination, see 14.7.1(4) Setting automatic startup and termination of the JP1/AJS3 Console services.
  3. Execute the jajs_setup_cluster command to set up the logical host, and create the shared files and directory on the shared disk.
    Make sure that the JP1/AJS3 service is not running on any logical and physical hosts, and then execute the jajs_setup_cluster command.
    The following shows the format of the jajs_setup_cluster command you execute:
     
    jajs_setup_cluster 
    [-h logical-host-name]
    -F scheduler-service-name
    -d shared-directory-name
    [-n scheduler-service-ID]
    [-m {cold|warm|hot}]
    {-P port-number-for-the-embedded-database
    -I embedded-database-setup-ID |
    -S
    }
     
    For details about the jajs_setup_cluster command, see jajs_setup_cluster in 2. Commands Used during Setup in the manual Job Management Partner 1/Automatic Job Management System 3 Command Reference 2.
    • For the -h option, specify the logical host name set in JP1/Base.
    • The -F option is required.
    • The -d option must be specified during setup of the primary node environment. For this option, specify a directory on the shared disk in which the shared directory and files will be created. As the shared directory, specified-directory-name/jp1ajs2/ is created, and the definition files on the local disk (files in /etc/opt/jp1ajs2/conf/) are copied to the directory that is created. Before you execute the command, make sure that the shared disk is mounted. If the -d option is omitted, the secondary node environment is set up.
    • If the -n option is omitted, the smallest available ID is assumed.
    • For the -m option, specify the scheduler service start mode. The specification of this option takes effect only during setup of the primary node environment. If this option is omitted, the value for the scheduler service specified in the DEFAULTSERVICENAME environment setting parameter is inherited.
    • Specify the -P and -I options only if you set up the embedded database when you set up the logical host.
      For the -P option, specify the port number for the embedded database used by the logical host.
      For the -I option, specify the setup ID of the embedded database used by the logical host.
      When the -P and -I options have been specified, the embedded database is set up with the following settings:
      - Database model: -s (small-scale model)
      - System logging: Not performed
      - System file duplication: Not performed
      - Unload log file: Not used
      - Data area directory: shared-directory/jp1ajs2/embdb\value-specified-for-the-I-option
      - Work area directory: /opt/jp1ajs2/embdb/value-specified-for-the-I-option/dbarea
      - Embedded database practical directory: /opt/jp1ajs2/embdb/value-specified-for-the-I-option
      For the other items, the default values are set.
    • Specify the -S option only if advanced setup is to be performed for the embedded database.
      If the -S option is specified, setup of the embedded database is skipped during setup of the logical host. Accordingly, the embedded database must be set up separately.
  4. Perform advanced setup for the embedded database.
    If you specified the -S option in step 3, perform advanced setup for the embedded database.
    For details about advanced setup for the embedded database, see D. Advanced Setup for the Embedded Database (in a Cluster Configuration).
  5. Specify the JP1/AJS3 Console Manager environment settings.
    If JP1/AJS3 Console Manager is being used, the JP1/AJS3 Console Manager environment must be set up for a cluster system.
    Note that you do not need to set up the JP1/AJS3 Console Agent environment for a cluster system because setup is performed during setup of the JP1/AJS3 Manager environment for cluster operation.
    Execute the following command to set the DATADIRECTORY environment setting parameter to the path to the data directory on the logical host:
     
    jajs_config -k [logical-host-name\JP1AJS2CONSOLEMANAGER] "DATADIRECTORY"="path-to-the-data-directory-on-the-logical-host"
     
    For example, if the logical host name is node0 and the logical host shared directory is /shdsk/node0, the command you execute is as follows:
    jajs_config -k [node0\JP1AJS2CONSOLEMANAGER] "DATADIRECTORY"="/shdsk/node0/jp1ajs2cm/database"
  6. Execute the following command to create the environment for executing jobs:
     
    jpqimport -dt isam -ci execution-environment-configuration-definition-file-for-queue-and-submit-jobs [-mh logical-host-name]
  7. To use queueless jobs, perform the necessary setup, as described below.
    Execute the following command:
     
    ajsqlsetup -h logical-host-name -F scheduler-service-name
     
    For details about the ajsqlsetup command, see ajsqlsetup in 3. Commands Used for Special Operation in the manual Job Management Partner 1/Automatic Job Management System 3 Command Reference 2.

    Cautionary note:
    Executing the ajsqlsetup command causes the files needed for queueless job execution to be created on the shared disk. Because the files are then checked, make sure that the shared disk is mounted before you attempt to execute this command.

This completes the tasks required on the primary node.

Cautionary notes:
  • Setup must be performed on each logical host.
  • When the logical host for the primary node is set up by using the jajs_setup_cluster command, the environment of the scheduler service specified in the DEFAULTSERVICENAME environment setting parameter is copied to create the logical host environment.
  • If the setting for starting multiple scheduler services is required on a logical host, see 16.2.7(1) Starting multiple scheduler services on a logical host.
  • When the logical host for the primary node is set up by using the jajs_setup_cluster command, the following files are copied from the physical host to the logical host:
    - jp1ajs_spmd.conf
    - jp1ajs_dbmd.conf
    - jp1ajs_hstd.conf
    - jp1ajs_schd.conf
    - jp1ajs_agtd.conf
    As a result, the settings on the physical host are inherited and processes that are not necessary for the logical host might be started. If the above files on the physical host have been customized, customize them again on the logical host so that only necessary processes are started.

(2) Tasks required on the secondary node (JP1/AJS3 - Manager)

To set up a cluster system environment for JP1/AJS3 - Manager on the secondary node:

  1. Set up the physical host.
    To do this, execute the following command:
     
    /opt/jp1ajs2/bin/jajs_setup
     
    If the physical host has already been set up, stop all JP1/AJS3 services, and then clear the scheduler service information from the shared memory of the system in the same way as on the primary node.
    If you use JP1/AJS3 Console Manager and JP1/AJS3 Console Agent, execute the following commands, respectively:
     
    /opt/jp1ajs2cm/bin/jp1ajs2cmsetup
    /opt/jp1ajs2/bin/jp1ajs2casetup
     
    You do not have to execute the above commands if the physical host has been set up. For the JP1/AJS3 Console service, automatic startup and termination must be set. For details about how to set automatic startup and termination, see 14.7.1(4) Setting automatic startup and termination of the JP1/AJS3 Console services.
  2. After completing the JP1/Base, JP1/AJS3, and JP1/IM tasks on the primary node, save the common definition information to the primary node and then set up the common definition information on the secondary node.
    After completing these tasks on the primary node, execute the jbsgetcnf command on the primary node to save the common definition information. Copy the backup file to the secondary node, specify the backup file as an argument of the jbssetcnf command, and then execute the command. The commands to be executed are as follows:

    On the primary node:
    jbsgetcnf -h logical-host-name > backup-file-name

    On the secondary node:
    jbssetcnf backup-file-name
  3. Perform the JP1/Base tasks required on the secondary node.
    For details about the JP1/Base tasks, see the Job Management Partner 1/Base User's Guide.
  4. Execute the jajs_setup_cluster command to set up the logical host.
    The following shows the format of the jajs_setup_cluster command you execute:
     
    jajs_setup_cluster
    [-h logical-host-name]
    -F scheduler-service-name
    [-S]
     
    • For the -h option, specify the logical host name set in JP1/Base.
    • For the -F option, always specify the scheduler service name that was specified on the primary node.
    • Specify the -S option only if advanced setup is performed for the embedded database.
      If the -S option is specified, setup of the embedded database is skipped during setup of the logical host. Accordingly, the embedded database must be set up separately.
  5. Perform advanced setup for the embedded database.
    If you specified the -S option in step 4, perform advanced setup for the embedded database.
    For details about advanced setup for the embedded database, see D. Advanced Setup for the Embedded Database (in a Cluster Configuration).
  6. To use queueless jobs, perform the necessary setup, as described below.
    Execute the following command:
     
    ajsqlsetup -h logical-host-name -F scheduler-service-name -nc
     
    For details about the ajsqlsetup command, see ajsqlsetup in 3. Commands Used for Special Operation in the manual Job Management Partner 1/Automatic Job Management System 3 Command Reference 2.

This completes the tasks required on the secondary node.

Cautionary notes:

(3) Tasks required on the primary node (JP1/AJS3 - Agent)

To set up a cluster system environment for JP1/AJS3 - Agent on the primary node:

  1. Perform the JP1/Base tasks required on the primary node.
    For details about the JP1/Base tasks, see the Job Management Partner 1/Base User's Guide.
    Note that, depending on the network configuration or operating environment, you might need to set the jp1hosts information. Specify the necessary settings by referring to the notes on cluster system operation and the procedure for setting the jp1hosts information in the Job Management Partner 1/Base User's Guide.
  2. Set up the physical host.
    For details about how to set up the physical host, see 12.2.2 Setting up JP1/AJS3 - Agent.
  3. Execute the jajs_setup_cluster command to set up the logical host, and create the shared files and directory on the shared disk.
    The following shows the format of the jajs_setup_cluster command you execute:
     
    jajs_setup_cluster -h logical-host-name [-d shared-directory-name]
     
    • For the -h option, specify the logical host name set in JP1/Base.
    • The -d option must be specified during setup of the primary node environment. For this option, specify a directory on the shared disk in which the shared directory and files will be created. As the shared directory, specified-directoryr-name/jp1ajs2/ is created, and the definition files on the local disk (files in /etc/opt/jp1ajs2/conf/) are copied to the directory that is created. Before you execute the command, make sure that the shared disk is mounted. If the -d option is omitted, the secondary node environment is set up.
  4. To use queueless jobs, perform the necessary setup, as described below.
    Execute the following command:
     
    ajsqlsetup -h logical-host-name
     
    For details about the ajsqlsetup command, see ajsqlsetup in 3. Commands Used for Special Operation in the manual Job Management Partner 1/Automatic Job Management System 3 Command Reference 2.

This completes the tasks required on the primary node.

Cautionary note:
Setup must be performed on each logical host.

(4) Tasks required on the secondary node (JP1/AJS3 - Agent)

To set up a cluster system environment for JP1/AJS3 - Agent on the secondary node:

  1. Set up the physical host.
    To do this, execute the following command:
     
    /opt/jp1ajs2/bin/jajs_setup
     
  2. After completing the JP1/Base, JP1/AJS3, and JP1/IM tasks on the primary node, save the common definition information to the primary node and then set up the common definition information on the secondary node.
    After completing these tasks on the primary node, execute the jbsgetcnf command on the primary node to save the common definition information. Copy the backup file to the secondary node, specify the backup file as an argument of the jbssetcnf command, and then execute the command. The commands to be executed are as follows:

    On the primary node:
    jbsgetcnf -h logical-host-name > backup-file-name

    On the secondary node:
    jbssetcnf backup-file-name
  3. Perform the JP1/Base tasks required on the secondary node.
    For details about the JP1/Base tasks, see the Job Management Partner 1/Base User's Guide.
  4. Execute the jajs_setup_cluster command to set up the logical host.
    The following shows the format of the jajs_setup_cluster command you execute:
     
    jajs_setup_cluster -h logical-host-name
     
    For the -h option, specify the logical host name set in JP1/Base.
  5. To use queueless jobs, perform the necessary setup, as described below.
    Execute the following command:
     
    ajsqlsetup -h logical-host-name
     
    For details about the ajsqlsetup command, see ajsqlsetup in 3. Commands Used for Special Operation in the manual Job Management Partner 1/Automatic Job Management System 3 Command Reference 2.

This completes the tasks required on the secondary node.

Cautionary note:
Setup must be performed on each logical host.

[Contents][Back][Next]


[Trademarks]

Copyright (C) 2009, 2010, Hitachi, Ltd.
Copyright (C) 2009, 2010, Hitachi Solutions, Ltd.