4.6.1 Example of execution for predictive error detection in the performance of monitored services and the corrective action support methodology
This subsection explains by way of example how to use SLM to execute predictive error detection in the performance of monitored services and the corrective actions to take, based on given conditions.
- Organization of this subsection
(1) Prerequisites
The conditions for this execution example are as follows:
Registration of monitored services and the setup required for predictive error detection have been completed and monitoring has already started.
The following figure shows the relationship among the personnel involved in this task.
Figure 4‒3: Relationship among personnel involved in predictive error detection in the performance of monitored services and the corrective action support methodology (execution example) Person who monitors all services
Instructs the monitor to perform monitoring. If notified of a warning sign of a service performance error, this person investigates the cause. Upon determining that further investigation is needed, this person asks the maintenance service provider for the monitored service to investigate.
Monitor
Uses the Home window to monitor the monitoring items for all monitored services that have been set up by the person who monitors all services.
Maintenance service provider for the monitored service
If requested by the person who monitors all services, this person investigates the monitored service and takes corrective action, as necessary.
(2) Predictive error detection in the performance of a monitored service
- Tasks in SLM
While the person who monitors all services was monitoring the status of the monitored services in the Home window, a warning constituting a warning sign of a service performance error was displayed.
The following figure shows a display example of the Home window when a warning is displayed for a monitored service.
Figure 4‒4: Display example of the Home window that contains a warning for a monitored service Details of the warning displayed in this figure are as follows:
-
When detected: 2020-02-05 06:41:03
-
Type: OUTLIER
-
Details: UPPER LIMIT
-
Service group: Group02
-
Service: Service02
-
Monitored target: All Web Access
-
Monitor item: Avg. response
This warning indicates that the average response time of Service02 belonging to Group02 that was obtained at 06:41:03 on February 5, 2020, constituted an out-of-range value (a value exceeding the upper limit) and differed significantly from the usual value for the monitored service.
-
- Results of the task
-
The monitor reported the warning to the person who monitors all services.
Because the warning might lead to an error if left unattended, the person who monitors all services decided to take corrective action immediately.
(3) Corrective action taken after a warning sign was detected in the performance of a monitored service
- Tasks in
After being notified of the warning displayed in the Home window, the person who monitors all services decided to use the Troubleshoot window to investigate the timing of the event detected as a warning, and then take corrective action.
The following figure shows a display example of the Troubleshoot window in which a warning is displayed for a monitored service.
Figure 4‒5: Display example of the Troubleshoot window in which a warning is displayed for a monitored service This performance chart of average response time indicates that the event causing the warning occurred between 06:39:03 and 06:43:03.
The access log for the time period during which the warning occurred include requests from the users of the Web system service and the corresponding responses. This information can be used to investigate any problems in Web system processing.
Figure 4‒6: Display example of an access log in which a warning for a monitored service is displayed - Results of tasks
Because the details of the warning and the timing of the event causing the warning became clear from the data provided in the Troubleshoot window, the person who monitors all services notified the maintenance service provider for the monitored service and requested a root cause investigation and corrective action.
(4) Verifying the service performance after taking corrective action
- Tasks in SLM
After corrective action was taken by the maintenance service provider for the monitored service based on the results of a root cause investigation, the person who monitors all services decided to use the Real-time Monitor window to verify that service performance had returned to normal.
The following figure shows a display example of the Real-time Monitor window showing that service performance has returned to normal after corrective action was taken.
Figure 4‒7: Display example of the Real-time Monitor window showing that service performance has returned to normal As shown in this figure, when service performance has returned to normal, the
(normal) icon is displayed in the Service performance information area.
- Results of tasks
The person who monitors all services has verified that service performance has returned to normal. This concludes the handling of the warning sign of a service performance error in a monitored service.