17.2.4 Registering the JP1/AJS3 service in the cluster software

Register the JP1/AJS3 service for the logical host in the cluster software that will be used. For details about how to register the service, see the documentation for the cluster software.

Note the following when registering the service:

Set the cluster software so that the secondary node can inherit the IP address and shared disk together with the JP1/AJS3 service from the primary node. In addition, set the cluster software so that the application program is also inherited.
Set the cluster software so that JP1/AJS3 starts after the secondary node has inherited the IP address and shared disk, and JP1/Base has started on the secondary node.
If you use JP1/AJS3 Console, you do not need to register the JP1/AJS3 Console service in the cluster software.

The following table describes the information that needs to be registered in the cluster software.

Table 17‒3: JP1/AJS3 information that needs to be registered in the cluster software
Function to be registered	Explanation
Start	Starts JP1/AJS3. Command used `jajs_start.cluster` Example `jajs_start.cluster` `logical-host-name` Time the start command ends The start command ends when JP1/AJS3 starts. However, if for some reason the startup processing is not completed within the timeout period (default: 60 seconds), the command ends midway. When the command ends, the startup processing is not canceled and still continues. Determining the result of the start command Use the operation monitoring function described below to determine the result of JP1/AJS3 startup. For details about the `jajs_start.cluster` command, see jajs_start.cluster (UNIX only) in 2. Commands in the manual Job Management Partner 1/Automatic Job Management System 3 Command Reference 1.
Stop	Stops JP1/AJS3. Command used `jajs_stop.cluster` Example `jajs_stop.cluster` `logical-host-name` Time the stop command ends The stop command ends when JP1/AJS3 stops. However, if for some reason the stop processing is not completed within the timeout period (default: 60 seconds), the command ends midway. When the command ends, the stop processing is not canceled and still continues. Determining the result of the stop command Use the operation monitoring function described below to determine the result of stopping JP1/AJS3. For details about the `jajs_stop.cluster` command, see jajs_stop.cluster (UNIX only) in 2. Commands in the manual Job Management Partner 1/Automatic Job Management System 3 Command Reference 1.
Operation monitoring	Monitors whether JP1/AJS3 is operating normally. Do not register this function if a failover is not required when a failure occurs in JP1/AJS3. Command used `jajs_spmd_status` Example `jajs_spmd_status -h` `logical-host-name` Determining the operation monitoring result The following describes the return values: Return value = 0 (normal) JP1/AJS3 is operating normally. Return value = 1 (error) An irrecoverable error has occurred. Assume that an error has occurred. Return value = 4 (partial stop) Some JP1/AJS3 processes have stopped for some reason. Assume that an error has occurred. Return value = 8 (all stop) All JP1/AJS3 processes have stopped for some reason. Assume that an error has occurred. Return value = 12 (status confirmation error) An error has occurred in the status confirmation processing of the `jajs_spmd_status` command. Retry checking of the operating status for the predefined number of times. For details about the `jajs_spmd_status` command, see jajs_spmd_status in 2. Commands in the manual Job Management Partner 1/Automatic Job Management System 3 Command Reference 1.
Forcible stop	Forcibly stops JP1/AJS3 and releases the resource being used. Command used `jajs_killall.cluster` Example `jajs_killall.cluster` `logical-host-name` Executing the `jajs_killall.cluster` command forcibly stops processes without performing any termination processing for JP1/AJS3. Execute this command only when a problem occurs (for example, processing does not stop when the stop command is executed). For details about the `jajs_killall.cluster` command, see jajs_killall.cluster (UNIX only) in 2. Commands in the manual Job Management Partner 1/Automatic Job Management System 3 Command Reference 1 and 11.4.1 Script for forcibly stopping JP1/AJS3 (jajs_killall.cluster) in the Job Management Partner 1/Automatic Job Management System 3 Administration Guide.

Setting the environment setting parameters

When the following environment setting parameter is set, the JP1/AJS3 services are stopped so that the cluster software can detect the error whenever the embedded database service is stopped due to an irrecoverable error.

Environment setting parameter	Explanation
`"RDBCHECKINTERVAL"=`	Checks the connection with the embedded database service, and automatically stops the scheduler service whenever disconnection is detected.

For details about these environment parameters, see 2.2 Setting up the scheduler service environment in the Job Management Partner 1/Automatic Job Management System 3 Configuration Guide 2.

For details about how to set up a scheduler service environment, see 14.2 Environment setting parameter settings and 17.2.3 Changing the common definition information.

To Page Top