In UNIX, when JavaVM terminates abnormally and a core dump is output, the information (stack trace information) required for investigating the cause for such abnormal termination can be acquired by the javatrace command. With the javatrace command, acquire the stack trace information from the core dump output. The javatrace command is installed in /opt/Cosminexus/jdk/jre/bin.

The execution format of the javatrace command is as follows:

 
javatrace core dump file name, Execution file name where core dump is created
 

When JavaVM terminates abnormally and you execute this command when a core dump is created with a file name called core, as a result, a file called javatrace.log is output under the current directory. Send this file to maintenance personnel.

Execution example

The example below describes a message that is output when JavaVM terminates abnormally and a core dump is created.

... # You can get further information from javatrace.log file generated # by using javatrace command. # usage: javatrace core-file-name loadmodule-name [out-file-name] [-l(library-name)...] # Please use javatrace command as follows and submit a bug report # to Hitachi with javatrace.log file: # [/opt/Cosminexus/jdk/jre/bin/javatrace core /opt/Cosminexus/CC/server/bin/cjstartsv] #

Execute the character string of the javatrace command displayed in the message. In the case of this example, execute /opt/Cosminexus/jdk/jre/bin/javatrace core /opt/Cosminexus/CC/server/bin/cjstartsv. As a result of execution, a file called javatrace.log is output under the current directory.

Note that the file name of the core dump actually output depends on the OS and it may be core.process ID in some cases. In such cases, specify the file name of the core dump actually output in the javatrace argument.