server aa...aa has failed. pid: bb...bb, service group name: cc...cc, run mode: dd...dd, critical state: e, ff...ff, end state: gg...gg

aa...aa: Failed server

bb...bb: Process ID of the failed server

cc...cc: Service group name of the failed server

dd...dd: Server run mode immediately before the server failed

OpenTP1: The server failed during execution of a library function provided by OpenTP1.

USER: The server failed during execution of something other than a library function provided by OpenTP1.

e: Indicates whether or not the server is in the critical status.

Y: The server is in the critical status.

N: The server is not in the critical status.

ff...ff: Critical information

gg...gg: Process end state returned with the UNIX system call wait (2)

In UNIX, the displayed process end state consists of four digits, where the upper two digits and lower two digits have different meanings.

The upper two digits indicate the value when the process was ended by exit(). For example, 0000 appears if the process was ended by exit(0), and 0100 appears if the process was ended by exit(1).

The lower two digits indicate the signal number when the process was ended by a signal. For example, 0009 indicates that the process was ended by SIGKILL (see /usr/include/sys/signal.h). The third digit of the process end state is 8 when a core file was output. For example, 008X indicates that the process was ended by signal number X and the core file was output.

The following gives examples of the process end state. Note that they are just examples, and other process end states may be displayed.

0083: Ended due to an OpenTP1 function (such as a timeout during real-time monitoring).

0086: Aborted.

008a: Ended due to an invalid memory access or other error.

008b: Ended due to an invalid memory access or other error.

0000: Ended by exit(0).

0100: Ended by exit(1).

****: The end state cannot be determined (by using, for example, the OpenTP1 dcsvstart command).

S: Performs postprocessing of the server and, if necessary, then restart it.

Countermeasure: If the run mode of the failed server is USER, the user server process ended in other than the OpenTP1 processing. Determine the user server processing that caused the server failure. If the run mode of the failed server is OpenTP1, determine the cause of the server failure and take corrective action according to the message that was output before this message. Then restart the server if necessary. If necessary, see the examples of errors and the investigation procedure in the manual OpenTP1 Operation.

For HP-UX:

When the user server failed, use the chatr command of the operating system to confirm that "immediate" is set for the bind mode.