1.1.1 Support for providing stable service

A service provider must be able to maintain the quality of the service and provide stable services to its users.

In other words, a service provider must set evaluation metrics (the service level objectives (SLOs)) to maintain the service level, and it must manage and run the service systematically.

To manage and run services systematically, it is helpful to apply a PDCA cycle. PDCA stands for Plan-Do-Check-Act, and ITSLM supports the tasks that correspond to C (Check) in the PDCA cycle.

The following figure shows the management and operation of services in a PDCA cycle when ITSLM is used.

Figure 1‒2: Management and operation of services in a PDCA cycle using ITSLM

Of the Check tasks, ITSLM supports service monitoring and evaluation. Service monitoring and evaluation by ITSLM involves performing the following tasks cyclically:

Define the configuration.

Define the services to be monitored.

ITSLM achieves the independence of each customer's business systems by using service groups to group monitored services by customer (such as a company) and setting access permissions required for monitoring each group. Therefore, before monitoring can be started, individual services must be registered into ITSLM and then their service groups must be defined.

ITSLM can help you register services by automatically detecting the URIs of the Web pages of the monitored services.
Set up monitoring.

Configure how to monitor the monitored services.

ITSLM specifies threshold values that will be used as evaluation metrics (SLOs) for maintaining the service level for each monitored service. Threshold values are provided for three items that are monitored: average response time, throughput, and error rate. Based on the specified threshold values, ITSLM can monitor for over-threshold values as well as possible future over-threshold values. The data obtained as a result of monitoring average response time, throughput, and error rate is referred to as the service performance. In addition to threshold values, you can also configure ITSLM to predict abnormalities in service performance stemming from unusual service statuses.
Monitor.

Monitor actual accesses to the services according to the monitoring settings.

ITSLM totals and analyzes actual accesses from service users and monitors for over-threshold values and possible future over-threshold values specified during monitoring setup, as well as for unusual service statuses (warning signs that might lead to abnormalities in service performance).
Evaluate periodically.

Output reports of accumulated daily service statuses as monitoring results.

Such reporting assists you in periodic evaluations to determine whether the evaluation metrics (SLOs) for maintaining the service level are being satisfied.

The monitoring task in 3 above requires some tasks that depend on the monitoring results. The following figure shows the tasks that must be performed depending on the monitoring task.

Figure 1‒3: Tasks that must be performed depending on the monitoring task

Monitoring, investigating the cause, and verifying recovery are performed in the cycle. Of the tasks that are performed as needed, ITSLM can support investigating the cause and verifying recovery.

Investigating the cause

If an abnormality or a warning sign that might lead to an abnormality is detected in the performance of a monitored service, its cause must be investigated promptly.

Because ITSLM can display ongoing service statuses (monitoring results) as graphs on the screen, the timing of an event that is the cause of a problem, or that might lead to a problem, can be identified more easily.

Verifying recovery

Because ITSLM monitors service statues, you can take an appropriate corrective action in response to a problem or a warning sign of an abnormality and then immediately check the current service status. This enables you to promptly determine whether services can be provided normally.

Thus, ITSLM plays an important role in the management and operation of services in the PDCA cycle and supports stable service operations.

To Page Top