JPS63276130A - Decentralized starting method for periodic task - Google Patents

Abstract

PURPOSE:To evenly divide a process in one and the same period by setting the different initial values at a timer which instructs the start of each periodic task and then setting successively these initial values under the time-out states in the order of smaller timer values. CONSTITUTION:A restoration monitoring means 11 monitors the restoration of a terminal controller 2 via a function block in a host computer 1 and informs the occurrence of an event to a periodic task starting means 12. An execution control means 13 for tasks contains the timers in the number corresponding to the cycle tasks, i.e., the number (n) of terminals. Then the means 12 gives a first timer setting request to the means 13 and the periodic tasks 141-14n give a 2nd and subsequent timer setting requests to the means 13. The information on the recovery is transmitted from the means 11 as a starting factor to the means 13.

Description

[Detailed Description of the Invention] [TR Required] A distributed activation method for periodic tasks processed by a computer, in which activation of periodic tasks is performed to avoid processing congestion when many periodic tasks are activated at the same time. The initial timer value of the timer that instructs is set to a different value for each task, so that the initial timeout is distributed over time.

[Industrial Field of Application] The present invention relates to a distributed activation method for periodic tasks used in a system including a host computer and a plurality of terminal devices connected to the host computer via a terminal control device.
More specifically, in the case where periodic processing (for example, failure detection processing) is performed on multiple terminal devices at the same time due to the occurrence of a certain event, we have developed a periodic task that prevents the processing from conflicting. This relates to a distributed startup method.

In a system where multiple terminal devices are connected to a host computer via a terminal control device and are configured to communicate, for example, communication between the host computer and the communication control device goes down, and then communication is restored. In such a case, the host computer takes the opportunity that communication is restored to control the terminal.
Processing is performed to start periodically transmitting test signals to each terminal device under the device, but in this case, if there are many terminal devices under the device, test signals from multiple devices are sent at once for each specified 1111. Therefore, a distributed startup method is required to reduce periodic processing congestion.

[Prior art] Conventionally, in this type of system, as a method to avoid periodic processing, (a) a method of varying the startup cycle for each task (b) a method of adjusting the startup time according to a startup time schedule There is.

Method (a) distributes the startup cycle by using a timer la that is shifted from the specified value for each task that performs periodic processing, and method (b) distributes the activation period for each task that performs periodic processing. The time in the cycle at which it should be started is entered in advance in the startup time schedule, and once the timer is set, it will be started according to this startup time schedule, and multiple tasks will be started at the same time. Activation cycles can be distributed by creating a startup timetable so that the

[Problems to be Solved by the Invention] However, the method (a) described above requires different startup cycles for each task, and therefore cannot be applied when deviation from the reference value is not allowed.

Furthermore, there is a problem in that the task activation frequency becomes uneven depending on the length of the activation cycle.

In addition, method (b) described above does not have the same problem as method (a), but the processing method is complicated because the startup time schedule must be managed in addition to the normal timing processing. There is also a problem that it is difficult to apply when the maximum number of periodic tasks is undefined.

The present invention has been made in view of these points, and
It is an object of the present invention to provide a method for distributing and starting periodic tasks that can effectively avoid retrieval of periodic processing.

Means for Solving Problem E] FIG. 1 is a flowchart showing the principle of the present invention. The present invention is an activation processing method in which a plurality of periodic tasks having the same period are simultaneously activated due to the occurrence of an event that becomes an activation factor, and a computer that executes the periodic tasks is Step 1) Set a different value for each periodic task as the initial timer value for the timer that instructs the activation of the timer. When the timer times out, start the target periodic task. Step 2). From the second time onwards, set a different value for each periodic task. Set the timer value of (step 3).

[Operation] Since different values are set as initial timer values in the timers instructing the activation of each periodic task, each timer times out in order from the one with the smallest timer value. Therefore, by distributing the processing evenly within the same cycle, it is possible to avoid periodic processing congestion.

[Example] Hereinafter, an example of the present invention will be described in detail with reference to the drawings.

FIG. 2 is a functional block diagram of a system implementing the method of the present invention. In the figure, 1 is a host computer, 2 is a terminal control device, and 31 to 1n are a plurality of terminal devices connected to the host computer W 1 via this terminal control device 2. In the following explanation, the periodic task processed by one host is the terminal group of each terminal device 1ff31 to 3n! ? Assume that the task is to perform monitoring.

In the In block in the host computer [11, 11 is a recovery monitoring means of the terminal control device 2, 12 is a periodic task starting means that is activated in response to an incident notification from the recovery monitoring means 11, and 13 is a host computer nva. It is an execution control means that controls the execution of tasks to be performed, and includes a timer corresponding to the number of periodic tasks (number of terminal devices). 141-14n
is a periodic task having the same cycle that is executed for each terminal device 31 to 3n, and is controlled by the execution control means 13.

Here, from the periodic task starting means 12 to the execution control means 13
An arrow $1 pointing toward indicates the flow of a signal instructing an initial timer set request to the periodic task 141;
shows the flow of a signal instructing an initial timer hit request to the periodic task 14n. Also, periodic task 14
Arrow 83.8 pointing from 1.14n to execution control means 13
4 indicates the flow of signals instructing the second and subsequent timer set requests of the periodic tasks 141 and 14n, respectively.

FIG. 3 is a block diagram showing the hardware of the system shown in FIG. Here, the terminal equipment r! shown in Fig. 2 is a switching system controlled by a computer. 131 to 3n are switchboards, and the terminal control device "eI2" is a switchboard control device.On the host computer 1 side, there are a digital switch module 59, a communication control device 6, a channel control wD device 7.
etc. are provided.

In the functional block diagram shown in FIG. 2, the periodic task activation means 12, which is the activation factor of each periodic task 141 to 14n, executes the restoration when a restoration notification is sent from the restoration monitoring means 11, for example. In response to the notification, a request is made to the execution control means 13 to start the periodic task. The execution control means 13 sets different initial timer values for the timers provided corresponding to the periodic tasks 141 to 14n, and distributes the initial activation times of the periodic tasks 141 to 14n. Also, from the second time onwards, the same timer value is set each time by the timer set request signal.

The above is an outline of the operation, and the details will be explained below with reference to the flowchart.

FIG. 4 is a flowchart showing the operation of the periodic task starting means 12. This operation flow is started when the terminal-based till device recovery monitoring and control means 11 detects communication recovery and receives a notification to that effect. First, it is determined whether a timer set request has been made for all of the terminals 141 to 14n (Step 11), and if not, the terminal! ! One of the i positions is selected and the initial timer value is calculated (step 12.13). This initial timer field is calculated by (terminal device number/total number of terminal devices) x specified timer value. Next, a timer set request (timer set acceptance call) is made to the execution control means 13 (step 14), and the process returns to step 11.

5 and 6 are flow charts showing the operation of the execution control means 13. ? - It is. First, the periodic task starting means 12
When a request for a timer is received from
The specified timer value is set in the area corresponding to the specified terminal device number according to the operation flow of >. Figure 5 (b
) is a conceptual diagram of areas corresponding to each terminal device number, and each area is configured to function as a timer provided for each terminal device.

FIG. 6 is an operational flowchart for monitoring timeouts of each timer. This operation flow always operates periodically with the unit time of the timer value, and first, it is determined whether the timer value has been decreased for all terminal devices whose timers are set (step 21), and if not, the timer value is Select one of the timers for the terminal device that has been set, subtract 1 from the timer value, and then determine whether the timer has timed out (is the timer value O?) (Steps 22, 23, 24>. If the timeout occurs, the target periodic task is started using the corresponding terminal device number as a parameter (step 25).

FIG. 7 is an operation flowchart of the periodic task activated in step 25 in FIG. Here, first, a test signal is sent to the corresponding terminal device (step 31), and then the terminal device and the prescribed timer value are designated and a timer set request for the second and subsequent times is made (step 32).

FIG. 8 is a timing chart of an embodiment that operates as described above.

A timer value is set in the timer corresponding to each terminal device (terminal device number is No, 1 to N00n) according to the operation flow shown in FIG. It was released in step 13 in
Terminal device number No. 1 is the smallest, and larger values are set in order, and terminal device number No. n is the largest value.

Therefore, as shown in FIG. 8, the timers corresponding to each terminal device are set in order from the timer corresponding to terminal device No. 1 to terminal device No. 1. No, 3. . . . Timeout occurs in the order of NO and n, and as a result, the corresponding periodic task 141, 1
42. ...14n will be activated in a distributed manner,
Periodic processing congestion is avoided.

For the second and subsequent timer values, the specified timer 11 is specified in step 32, and thereafter the effect of processing distribution is maintained, and the activation frequency is not uniform and does not deviate from the reference value. There is.

In the above embodiment, communication recovery of the terminal control device is taken as an example of the event that causes the periodic task to start #J, and
Although the periodic task is assumed to be a terminal failure monitoring task such as sending out a test signal, it is not limited to this.

[Effects of the Invention] As explained in detail above, according to the present invention, by setting different initial timer values to the timers that instruct each terminal device to start a periodic task, periodic tasks can be performed. Clearance of processing can be effectively avoided.

[Brief explanation of the drawing]

Fig. 1 is a flowchart showing the principle of the present invention, Fig. 2 is a functional block diagram of a system that implements the method of the invention, Fig. 3 is a configuration block diagram showing the hardware of the system shown in Fig. 2, and Fig. 4 is a block diagram showing the hardware of the system. FIG. 2 is a flowchart showing the operation of the periodic task starting means; FIGS. 5 and 6 are flowcharts showing the operation of the execution control means; FIG. 7 is a flowchart showing the operation of the periodic task; FIG. This is a timing chart. In FIG. 2, 1 is a host computer, 2 is a terminal control device, 31 to 3n are terminal devices, 11 is a recovery monitoring stage 1, 12 is a periodic task starting means, 13 is an execution control means, and 141 to 14n are periodic tasks. It is. Flowchart for adding uninvented principles Fig. 1 Flowchart for periodic task activation means Fig. 4 Fig. 6

Claims (1)

[Scope of Claims] A startup processing method in a computer in which a plurality of periodic tasks having the same period are activated at the same time and the periodic tasks are executed due to the occurrence of an event that becomes a activation factor, the method comprising: instructing the activation of each of the periodic tasks; Set a different value for each periodic task as the initial timer value in the timer that is used (Step 1), and when the timer times out, start the target periodic task (Step 2), and from the second time onwards, set a different value for each periodic task as the initial timer value. Set the timer value (
Step 3) A distributed activation method for periodic tasks, characterized in that: