3.1 Monitoring supported by ITSLM
ITSLM monitors the following three monitoring items based on actual accesses from users to the monitored services:
-
Average response time
-
Throughput
-
Error rate
The data obtained by monitoring the monitoring items (average response time, throughput, and error rate) is characterized as the service performance. Service performance represents one second's worth of data, which means that service performance is measured 60 times per minute.
ITSLM enables you to perform out-of-range value detection and SLO monitoring on the basis of the monitoring items. The following table describes out-of-range value detection and SLO monitoring.
No. |
Monitoring (detection) type |
Description |
|
---|---|---|---|
1 |
Out-of-range value detection |
If the performance of a monitored service varies significantly from what is typical, this monitoring method regards such a condition as an early warning sign of a potential service performance error. |
|
2 |
SLO monitoring |
Trend monitoring |
This monitoring method determines trends in the performance of a monitored service and uses the trends to predict overages of a service performance threshold. |
Threshold value monitoring |
This monitoring method detects an overage of a service performance threshold for a monitored service. |
When out-of-range value detection, trend monitoring, and threshold value monitoring are all performed, a warning is displayed by out-of-range value detection and trend monitoring whenever the possibility of a service performance error in a monitored service is suspected. If you take an appropriate corrective action at this early stage, you can prevent the service performance error from occurring. Once a service performance error has occurred, it is displayed by threshold value monitoring. In such a case, immediate corrective action is assumed to be called for.
If you link ITSLM with Performance Management, you can monitor the hosts and middleware that provide the monitored services. The monitoring items are set in Performance Management beforehand. The performance data monitored by Performance Management is called system performance. Out-of-range value detection and SLO monitoring are also applicable to system performance, similarly to service performance.
In addition, by linking ITSLM with Performance Management, you can monitor the availability of monitored services. Availability monitoring detects monitored services that have stopped as a result of an error. You can obtain the following evaluation metrics (SLOs) on the basis of the availability information acquired by availability monitoring:
-
Service availability
-
Mean time to recovery
-
Mean time between failures
- Organization of this section
-
-
3.1.1 ITSLM's monitoring methods and types of monitored targets
-
3.1.2 Using out-of-range value detection for detection of unusual status in monitored services
-
3.1.3 Using trend monitoring for detection in advance of threshold overages
-
3.1.4 Using threshold value monitoring for detection of threshold overages
-