Appendix F.1 Flow of setting up a system

For editing files and executing commands to set up a system for executing batch applications, you use commands and user-defined files provided with the Application Server and the operating environment of the system.

Organization of this subsection
(1) Files used when setting up the system
(2) Flow of setting up a system that executes batch applications
(3) Data required for troubleshooting

(1) Files used when setting up the system

This subsection describes the main files used when setting up a system that executes batch applications. For details on each file, see 3. Files Used on Batch Servers in the uCosminexus Application Server Definition Reference Guide.

Note
To change a file, stop the batch server. The operation will not be guaranteed if you change the contents of a file while the batch server is running.

(2) Flow of setting up a system that executes batch applications

To set up a system that executes batch applications by editing files and executing commands:

  1. Set up a batch server.
    For details, see F.2 Setting up batch servers.
  2. Specify the operation settings of the batch server.
    When changing operations of the functionality to be used or using the specific functionality, you use user-defined files and commands that are provided with the Application Server for setting up or changing operations of the batch server. However, you make sure that the transaction is set to the local transaction.
    The functionality for which you can set up or change the operation of the batch server are as follows:
    • Naming Service
    • JavaVM startup options
    • Container extension library
    For the settings of the main functionality provided with the Application Server, see 2.3.10 Settings in the execution environment (Batch server settings) in the uCosminexus Application Server Expansion Guide.
  3. Specify the settings for using the job scheduling functionality.
    These settings are required only for scheduling the execution of batch applications. The scheduling functionality uses CTM. You specify the operation settings for using CTM in the following files:
    • The ejbserver.ctm.enabled property in usrconf.properties (user property file for batch servers)
      For details, see 3.3 usrconf.properties (User property file for batch servers) in the uCosminexus Application Server Definition Reference Guide.
    • The batch.ctm.enabled option in usrconf.cfg (option definition file for batch applications) for the commands (cjexecjob, cjkilljob, and cjlistjob) used in batch applications
      For details, see 3.6 usrconf.cfg (Option definition file for batch applications) in the uCosminexus Application Server Definition Reference Guide.
  4. Specify the settings for collecting data on troubleshooting.
    If you are using the default settings, you need not specify the settings. However, if you do not specify the settings beforehand for collecting the following data, you will not be able to collect data. This data is required for troubleshooting, and therefore, we recommend that you collect this data.
    • Statistics of the OS, user dump (in Windows), or core dump (in UNIX)
    • Log of the JavaVM garbage collection
    When collecting this data, or when changing the settings for collecting some other data, specify the settings for collecting the troubleshooting data. For details on data required for troubleshooting, or whether or not specifying the settings for troubleshooting is required, see F.1(3) Data required for troubleshooting.
  5. Specify the settings for resources.
    Use server management commands to define the properties of resource adapters, and import them into the batch server. For details on specifying the settings, see 2.7.6 Setting flow of resource adapter in the uCosminexus Application Server Expansion Guide.
  6. Start the system, and check the system operation.
    For details on starting and stopping the system, see G Starting and Stopping a System that Executes Batch Applications (in Windows), or H Starting and Stopping a System that Executes Batch Applications (in UNIX).

(3) Data required for troubleshooting

The following table describes data required for troubleshooting, and whether or not specifying the settings for troubleshooting is required. For details on specifying the settings, see 3. Preparation for Troubleshooting in the uCosminexus Application Server Maintenance and Migration Guide.

Table F-1 Types of troubleshooting data and whether or not specifying the settings (for a system that executes batch applications)

Type of dataPossibility of specifying the settings
snapshot log[Figure]#1
Trace based performance analysis fileN#2
Batch server log[Figure]
Application user logY
Statistics file[Figure]
Resource adapter log[Figure]
Trace file of Cosminexus TPBroker[Figure]
Log of Cosminexus DABroker Library[Figure]
Statistical information of the OSY
User dumpY
core dumpY
JavaVM dataY
Log of server management commands[Figure]
Legend:
Y: Settings must be specified when collecting data.
[Figure]: Settings must be specified only when changing the default settings for collecting data.
N: Settings need not be specified.
#1
Collect the snapshot logs using the snapshotlog command. In such a case, during the execution of the command, specify the already set up list files as the arguments. For details on the snapshotlog command, and the snapshot logs that can be collected with the default settings, see snapshotlog (collect snapshot logs) of 11.2 Details of commands used in log operations. To change the snapshot log collection destination, edit the snapshot log collection target definition files. For the snapshot log collection target definition files, see 12.2 Definition file for collecting snapshot log in the uCosminexus Application Server Definition Reference Guide.
#2
During daily system operations, when you start the performance tracer by executing the cprfstart command, you can specify the PRF trace collection level and the number of files. For collecting the trace based performance analysis file, see 7.3.1 How to collect the trace based performance analysis files in the uCosminexus Application Server Maintenance and Migration Guide.

The logs described in the table can be collected in a batch as the snapshot log. However, some trace files of Cosminexus TPBroker can be collected, while some cannot be collected. Besides this, you must specify the settings for collecting the log that cannot be collected with the default settings of the snapshot log, and add them to the snapshot log collection destination.

Note that you cannot change the log output destination of the following logs:

For the types and default values of logs, channel names, details of the logs that can be collected and collection methods, see 2.4 Types of data that must be collected in the uCosminexus Application Server Maintenance and Migration Guide.