17.5 Data to be collected in the event of trouble
If the actions described in 17.2 Troubleshooting are not successful in correcting the error, collect the necessary data, and then contact the system administrator to determine the cause of the error. This subsection describes the data that you need to collect in the event of an error.
Performance Management provides commands for collecting the needed data in one operation. To collect PFM - Manager, PFM - Agent, and PFM - RM data, use the jpcras command. To collect PFM - Web Console data, use the jpcwras command. The following tables indicate the data that can be collected by the jpcras or jpcwras command.
- Note:
-
The data collected by the jpcras or jpcwras command depends on the options you specify when you execute the command. For details on the options specified in the command and the data that can be collected, see the chapter explaining the commands in the manual JP1/Performance Management Reference.
- Notes on logical host use:
-
Note the following when using a logical host.
-
The logs for Performance Management for logical host use are stored on the shared disk. When the shared disk is online (in Windows) or mounted (in UNIX), the jpcras command can be used to collect all logs on the shared disk at once.
-
To determine the cause of any problems at the time of a failover, the data before and after the failover is required. Therefore, the data of both the executing and standby nodes is required.
-
To investigate Performance Management for logical host use, the cluster software data is required. Because the starting and stopping of Performance Management for logical host use are controlled by the cluster software, investigate Performance Management by comparing the operations of the cluster software and Performance Management.
-