JPS58151743A - Polling system - Google Patents

Abstract

PURPOSE:To ensure the uniform services to all stations, by changing a station address with which the polling is started for each polling period. CONSTITUTION:The slave work stations 1-8 are connected to a master work station 10 via a serial interface connecting cable 9. A processing buffer 11 is provided within the station 10 to accept an input request. The buffer 11 is shared by all stations. The queuing cases for the input request are increased for the station that receives services in the latter half of the polling. Therefore the polling is started from an SS8 for the 2nd polling period in case the polling is carried out in order of SS1-SS8 in the 1st polling period. Thus other stations are delayed by an order respectively. Hereafter the order is changed in the same way until the 8th polling in order to ensure the uniform services.

Description

DETAILED DESCRIPTION OF THE INVENTION The invention relates to a polling method, and particularly to a method for equalizing polling when a master workstation controls multiple slave workstations using a multi-drop serial interface control method. Concerning the polling method to be done on the Cape.

Prior Art In the conventional L-ring system, the polling operation is repeated in order from the lowest number of stage address. A fixed polling method is common.

However, after polling all stations in order in one polling cycle, an interval is taken to process input requests, etc., and the same operation is repeated again in the next polling cycle. For systems that do this, the above-mentioned progress is made.

According to a system in which the polling order of stations is fixed, as shown in FIG.

In particular, in systems where there is a limit to the capacity of the processing buffer that accepts input requests, a situation may arise in which the processing buffer temporarily becomes free, and in this case, the waiting time increases and becomes extremely In order to solve this problem, it is possible to always prepare a large number of processing buffers for stations, but this has another problem, which is that memory usage becomes very inefficient. This will cause problems. Additionally, in some cases, systems may not be able to provide such a large enough processing buffer.
In reality, there are many cases where this is impossible.

Purpose of the Invention The present invention has been made in view of the above circumstances, and its purpose is to replace the conventional polling method.

The object of the present invention is to provide a polling method that solves the above-mentioned problems and mills services for all stations.

The key point of the present invention is to change the station address at which polling is started every polling cycle (time cycle of polling all stations in order). The point is that the station that received service later in the cycle is controlled by software so that it receives service earlier in the next polling period, thereby equalizing the service for all stages.

Embodiments of the present invention will be described in detail below with reference to the drawings.

Embodiment of the invention and its effects FIG. 1 is a four-point diagram showing an embodiment of the invention. In this embodiment, eight slave workstations 1 to 8 are connected to a master workstation 10 by a serial interface connection cable 9. A processing buffer 11 for receiving input requests is provided within the master station IO.

FIG. 2 is a diagram showing the polling order that changes for each polling cycle for slave workstations, and 881 to 888 are slave workstations 1
It corresponds to 8.

In the process of master workstation 1o polling slave workstations 1 to 8,
When an input request occurs, the request is taken into the processing buffer 11, and after the processing is completed, the processing buffer 11 is released in preparation for the next request. Since the processing buffer 11 is configured to be shared by all stations, input requests are often made to wait for stage stations that receive service in the latter half of polling. Therefore,
In this embodiment, as shown in FIG. 2, the station-based services are changed such that the station that received the service last in the current polling cycle will receive the service first in the next polling cycle. ,
Service to all stations is now equalized in units of polling cycles equal to the number of stages.

As shown in FIG. 2, in #61 polling cycles, S S l -! - If polling is performed in the order of S S 8, in the second polling cycle, polling will start from the SSB that was polled last in the first cycle, and other stations will be polled in the same order as in the first cycle. Each is passed one step later. Thereafter, the order is changed in the same way up to the 8th time, and services to all stations are equalized with 8 times being one unit.

FIG. 3 shows the program processing of seven stomachs and chards in the apparatus of this embodiment. As is clear from FIG. 3, one polling cycle is a polling cycle.

After one cycle starting from the slave stage indicated by the pull start pointer, the start pointer of the polling table is updated (in this case, it is updated to the last polled station address) and the process ends.

Note that in the above embodiment, the station that last received service in the first polling cycle is
! Although an example has been shown in which the rankings change one by one, such as being the first station to receive service in a bowling mu, the number of stations whose rankings are changed is not limited to one.

Further, in the above embodiment, an example was shown in which the change of station service order is controlled by four-gram control, but this can also be done by a combination of logic circuits, and the present invention is applicable to such an implementation. It is clear that the term also includes aspects.

Effects of the Invention As described above, according to the present invention, in the polling method of a workstation system in which one master workstation controls a plurality of slave workstations using the serial interface l11m method, Since stations that received service later in the polling cycle are controlled to be polled earlier in the next polling cycle, service to all stations is equalized.
, while achieving the average performance of all stations,
This has the remarkable effect that it is possible to share a processing buffer using the minimum necessary memory capacity.

[Brief explanation of drawings]

FIG. 1 shows a workstation illustrating an embodiment of the present invention.
FIG. 2 is a block diagram of the system, FIG. 2 is a diagram showing the polling order that changes for each polling cycle for slave workstations, and FIG. 3 is a processing flowchart. 1 to ♂ Nislepe workstation, 9 real interface connection cable, lO: master workstation, 118 processing buffer, g1 to 24: processing step, 25: polling fist table, 26: start pointer. Patent applicant Kinsha Co., Ltd. Hitachi, Ltd. No. 11゛〈 Fig. 2

Claims (2)

[Claims] (1) In a polling method that controls one master workstage and multiple slave workstations using a serial interface control method,
By updating the pointer of the polling table provided corresponding to the plurality of slave workstations, polling of the plurality of slave workstations is performed in a certain polling cycle, and the polling is performed one after another. This method is characterized in that the slave fist work stage which has been serviced is controlled so that it is serviced in the previous position in the next polling cycle. (2) The polling of the slave workstations is repeated in a certain polling cycle, and the slave workstation that was serviced at the end is placed in the next polling cycle, and then the slave workstation is serviced at the top. A polling method according to claim S, characterized in that the polling method is controlled as follows.