3.1.5 Process monitoring examples
This subsection explains how to monitor process performance.
- Organization of this subsection
(1) Overview
Because system functionality is provided by individual processes, understanding the operating status of processes is essential for stable system operation.
If one of the processes that provides system functionality terminates abnormally, the system stops with serious consequences. In order to detect such an abnormal condition early and take appropriate action, it is necessary to monitor the status of processes, including their generation and disappearance.
Note that PFM - Agent for Platform performs a process check at the same intervals that information is collected. Accordingly, the time that the disappearance of a process is detected is the time that PFM - Agent for Platform collects information, not the actual time that the process disappeared.
The following table lists and describes the principal records and fields related to the monitoring of processes.
|
Record |
Field |
Description (example) |
|---|---|---|
|
PI_WGRP |
Process Count |
The number of processes. If the value of this field is the threshold or less (the minimum number of processes that need to be activated), some or all of the required processes are not running.# |
|
PD_PDI |
Program |
The name of a process. If this record is not collected, the process is not running. |
|
PD_APS |
Program Name |
The name of a process. If this record is not collected, the process has stopped. |
|
PD_APP, PD_APP2 |
Application Name |
The name of an application definition |
|
Application Exist |
A status of the applications. NORMAL indicates that the status of any one of the monitored targets is NORMAL. ABNORMAL indicates that the status of all the monitored targets is ABNORMAL. |
|
|
Application Status |
A status of the applications. NORMAL indicates that the status of all the monitored targets is NORMAL. ABNORMAL indicates that the status of any one of the monitored targets is ABNORMAL. |
|
|
PD_APPD |
Application Name |
Conditional results on the number of monitors. If the value of the Monitoring Status field is ABNORMAL, the number of running programs or command lines is not within the specified range. |
|
Monitoring Label |
||
|
Monitoring Status |
- #
-
The /opt/jp1pc/agtu/agent/wgfile file must be setup to collect this record.
(2) Monitoring methods
(a) Monitoring process disappearance
You can use the Process End alarm provided by the monitoring template to monitor process disappearance.
If a process terminates abnormally, the system stops with serious consequences. You can monitor the disappearance of processes by using an alarm, enabling prompt recovery of the system.
For details, see 3.2.5(1) Monitoring template.
(b) Monitoring process generation
You can use the Process Alive alarm provided by the monitoring template to monitor process generation.
You can use an alarm to monitor the generation of processes for each application or the status of scheduled processes, enabling you to check the operation status of the production system.
By setting a workgroup in the wgfile file and using the PI_WGRP record, you can perform several types of monitoring. For example, you can monitor the following items: process generation, process disappearance, the number of processes that have the same name, the number of processes for each application, and the number of processes activated for each user.
For details, see 3.2.5(1) Monitoring template.