18.5.4 Errors detected by the health check function

The health check function can detect errors in the JP1/Base processes. The following describes the causes and recovery actions for errors detected by the health check function.

Organization of this subsection

(1) There is a large number of system resources (CPU, disk, and other resources) being consumed. Or, the number of process requests exceeds the performance limit.
(2) The command process does not end as expected. Or, the command process does not end and still retains system resources.
(3) A process is in a deadlock or infinite loop
(4) Unable to connect to the host to be monitored

(1) There is a large number of system resources (CPU, disk, and other resources) being consumed. Or, the number of process requests exceeds the performance limit.

Cancel any processing that places a high load on the system.

To Page Top

(2) The command process does not end as expected. Or, the command process does not end and still retains system resources.

Using an OS function such as the kill command, forcibly end the command process.

To Page Top

(3) A process is in a deadlock or infinite loop

If a process goes into a deadlock or infinite loop and fails to end in a timely manner, take the recovery action described in the following table.

No.	Function	Process name		Recovery action
1	Process management	`jbs_spmd`		Restart JP1/Base. In Windows: Restart the JP1/Base services (process management including user management). In UNIX: Restart JP1/Base.^#
2	Authentication server	`jbssessionmgr`
3	Configuration management	`jbsroute`
4	Command execution	`jcocmd`
5	Plugin service	`jbsplugind`
6	Event service	`jevservice`		Restart the event service. In Windows: Restart the JP1/Base Event service. In UNIX: Restart the event service.^#
7	Log file trap	`jevtraplog`	`jevtraplog`	Restart the log-file trap management service (daemon). In Windows: Restart the JP1/Base LogTrap service. In UNIX: Restart the log-file trap management daemon.^#
			`jevlogd`
			`jelparentim`	Using the `jevlogstart` command, restart the log file trap that has the ID indicated in the error message. In Windows: Restart the log file trap whose thread ID is the same as the ID indicated in the message. In UNIX: Restart the log file trap whose process ID is the same as the ID indicated in the message.
			`jelchildim`
8	Event log trap	`jevtrapevt`		Restart the event log trapping service (JP1/Base EventlogTrap).
9	SNMP trap converter	`imevtgw`		Restart NNM.
10	Health check	`jbshcd`		Restart JP1/Base. In Windows: Restart the JP1/Base services (process management including user management). In UNIX: Restart JP1/Base.^#
10	Health check	`jbshchostd`
11	Service management control	`jbssrvmgr`
12	Local action	`jbslcact`
13	Inter-process communication	`jbscomd`

#: After terminating the processes with the stop command, use the ps -el command to make sure all the processes have ended. If any processes are still active, end them with the kill command. Then restart the processes using the start command.

To Page Top

(4) Unable to connect to the host to be monitored

Check whether the host has started.
Check whether JP1/Base has started.
Check whether there is a problem on the network.
Make sure that JP1/Base installed on the host to be monitored is version 07-51 or later.

To Page Top