3.9.2 Example of monitoring a schedule queue

This subsection provides an example to explain the monitoring of a schedule queue.

The example in this subsection assumes that the following settings have been specified:

• Queue filling percentage triggering schedule queue monitoring: 60%

• Threshold for stopping the system: 70%

• Schedule queue check interval: 1 second

  Figure 3‒22: Example of schedule queue monitoring

In this example, the threshold of the system processing percentage for stopping the system is set to 70%. Therefore, the right side of the schedule queue monitoring expression (M1 / 100) is 70 / 100 (= 0.7). Therefore, the expression in this example is as follows:

Schedule queue monitoring expression in this example:

(P / C_{n - 1}) < 0.7

In this case, the CTM daemon terminates abnormally when the left side (P / C_{n - 1}) is less than 0.7.

This example also assumes that the maximum number of requests that can be contained in the schedule queue is 50. Therefore, 60% of schedule queue filling percentage means 30 queued requests. Therefore, monitoring of the schedule queue starts when the number of queued requests exceeds 30.

The following describes the processing of schedule queue monitoring for each check point:

C₁

At check point C₁, 31 requests are in the schedule queue. Because the schedule queue filling percentage exceeds 60% (or, 30 requests), monitoring of the schedule queue starts.

C₂

At check point C₂, P (number of requests processed from C₁ to C₂) = 22. Therefore, the left side of the schedule queue monitoring expression (P / C_{n - 1}) is as follows:

(P / C₁) = 22 / 31 = 0.7

Because the resulting value (0.7) is equal to the threshold for stopping the system (70%), the CTM daemon does not stop.

Also, the number of queued requests at C₂ is 45. Because this number exceeds the queue filling percentage triggering schedule queue monitoring, 60% (or, 30 requests), monitoring of the schedule queue continues.

C₃

At check point C₃, P (number of requests processed from C₂ to C₃) = 32. Therefore, the left side of the schedule queue monitoring expression (P / C_{n - 1}) is as follows:

(P / C₂) = 32 / 45 = 0.71

Because the resulting value (0.71) is greater than the threshold for stopping the system (70%), the CTM daemon does not stop.

Also, the number of queued requests at C₃ is 35. Because this number exceeds the queue filling percentage triggering schedule queue monitoring, 60% (or, 30 requests), monitoring of the schedule queue continues.

C₄

At check point C₄, P (number of requests processed from C₃ to C₄) = 35. Therefore, the left side of the schedule queue monitoring expression (P / C_{n - 1}) is as follows:

(P / C₃) = 35 / 35 = 1

Because the resulting value (1) is greater than the threshold for stopping the system (70%), the CTM daemon does not stop.

Also, the number of queued requests at C₄ is 30. Because this number is equal to the queue filling percentage triggering schedule queue monitoring, 60% (or, 30 requests), monitoring of the schedule queue stops.

C₅

At check point C₅, P (number of requests processed from C₄ to C₅) = 5. Therefore, the left side of the schedule queue monitoring expression (P / C_{n - 1}) is as follows:

(P / C₄) = 5 / 30 = 0.16

Although this value is less than 0.7 (70%), the threshold for stopping the system, the CTM daemon does not stop because the schedule queue is not monitored at C₅.

C₆

At check point C₆, 31 requests are in the schedule queue. Because the queue filling percentage triggering schedule queue monitoring, 60% (or, 30 requests), is exceeded, monitoring of the schedule queue starts.

C₇

At check point C₇, P (number of requests processed from C₆ to C₇) = 2. Therefore, the left side of the schedule queue monitoring expression (P / C_{n - 1}) is as follows:

(P / C₆) = 2 / 31 = 0.06

Because this value is less than 0.7 (70%), the threshold for stopping the system, the CTM daemon terminates abnormally.

To Page Top