Job Management Partner 1/Integrated Management - Manager Overview and System Design Guide

12.7.2 Troubleshooting in JP1/IM and JP1/Base

System operation management is affected by any failure that causes JP1/IM or JP1/Base to stop. JP1/IM and JP1/Base therefore provides the following functionality to enhance failure tolerance:

• Automatic restart if a process ends abnormally (process management)
• Issuing of a JP1 event when an error is detected during process start or stop processing (process management)
• Issuing of a JP1 event and execution of a notification command when a hangup is detected in a process (health check function)

Consider enabling these functions. They are disabled by default.

About process error troubleshooting and reporting:

• Process management
  See 7.1 JP1/IM - Manager process management.
  
• Setting to enable process restart after an error
  See Extended startup process definition file (jp1co_service.conf) in 2. Definition Files in the manual Job Management Partner 1/Integrated Management - Manager Command and Definition File Reference.
  
• Setting to enable issuing of a JP1 event when a process error occurs
  See IM parameter definition file (jp1co_param_V7.conf)"> in 2. Definition Files in the manual Job Management Partner 1/Integrated Management - Manager Command and Definition File Reference.
  
• Health check function
  See 7.2 JP1/IM - Manager health check function.
  
• Setting to enable issuing of a JP1 event and execution of a notification command on detection of a process hangup
  See Health check definition file (jcohc.conf)"> in 2. Definition Files in the manual Job Management Partner 1/Integrated Management - Manager Command and Definition File Reference.
  
• JP1/Base troubleshooting and error reporting
  See the description of setup for troubleshooting JP1/Base errors in the Job Management Partner 1/Base User's Guide.
  

To enable prompt retrieval of relevant information in the event of a failure, JP1/IM provides dump output commands and data collection tools. These are normally used to collect data when a problem occurs in JP1/IM. However, they cannot be used with the Web-based JP1/IM - View; instead, you must use the debugging tools provided by Java^TM Plug-in to collect error information.

Dump output commands

JP1/IM - View and JP1/IM - Manager each have their own dump output command.

Table 12-9 Dump output commands

Command	Description
jcothreaddmp	Dump output command for JP1/IM - View. Outputs the following dump as diagnostic data when a hangup occurs in JP1/IM - View: Java thread dump

Data collection tools

The data collection tools are provided as a batch files (in Windows) or scripts (in UNIX).

Table 12-10 Data collection tools provided by JP1/IM

OS	Data collection tool	Description
Windows	jcoview_log.bat	Tool for collecting in a batch all data required for error investigation in JP1/IM - View
	jim_log.bat	Tool for collecting in a batch all data required for error investigation in JP1/IM - Manager#
UNIX	jim_log.sh	Tool for collecting in a batch all data required for error investigation in JP1/IM - Manager#

#: To collect data for investigating an error in JP1/Base, you must execute the data collection tool provided by JP1/Base. (The JP1/Base data collected by the tools in the above table relates to JP1/IM operation only.)

For details about the collected data, see 9.3 Data that needs to be collected when a problem occurs in the Job Management Partner 1/Integrated Management - Manager Administration Guide. In Windows, a memory dump or crash dump may be required if a problem occurs. We recommend that you set Windows to output these dump files in case they are needed.

However, note the following points:

• The size of the memory dump depends on the real memory size. The larger the installed physical memory, the larger the memory dump. Allocate sufficient disk space to collect a memory dump. For details, see STOP errors in the Windows help.
• Not only JP1 event information but also error data for other application programs is output in a crash dump. Output of a crash dump reduces the amount of available disk space by the volume of output data. Allocate sufficient disk space if you have set Windows to output a crash dump.

Java^TM Plug-in debugging tool

If a hangup occurs while you are using the Web-based JP1/IM - View, collect a Java stack trace log using the debugging tool provided by Java^TM Plug-in. You must first set up the Java^TM Plug-in Control Panel window so that the tool will be accessible when the Web-based JP1/IM - View is running. For details, see 4.14.4 Specifying display settings for the Java Console window in the Job Management Partner 1/Integrated Management - Manager Configuration Guide.

[Contents][Back][Next]

All Rights Reserved. Copyright (C) 2009, Hitachi, Ltd.