JPH0666807B2 - Loop communication system - Google Patents

Description

The present invention relates to a loop communication system for performing data communication between a plurality of computers, and more particularly to a token ring type loop communication system.

[Conventional technology]

2. Description of the Related Art Conventionally, a device that connects computer systems to each other via their own communication control devices and loop communication lines to perform data communication by the token ring system is known, for example, from Japanese Patent Laid-Open No. 60-160747. There is.

In this type of loop communication system, the communication control devices of a plurality of computers to be subjected to data communication are connected to each other via a loop communication line, and in order to control token issue management, One is the token issue management station.

Then, the issued token is placed on the loop communication line and circulated from the token issuing management station to each communication control device in the order of connection, and when the circulated token reaches each communication control device, the text (computer data) If there is a request to send the token, the token issuance name is converted to its own station name, the text is sent, and the token is sent to the next communication control unit.
If there is no request for text transmission, the token is sent as it is to the next communication control unit. If there is a text transmission request from two or more communication control devices in sequence during the token circulation, the transmission text is sequentially concatenated until the text is received by the destination communication control device, and then the token (in this case, The token will be rewritten every time there is a transmission request, so as a result, it will be the name of the communication control device of the station that made the last text transmission request). It is sent to the communication control device.

In the process of circulating the tokens in order, if there is no text transmission request in any of the communication control devices, the tokens return to the token issue management station in the original name of the original station. In this case, the issuing management station once discards (erases) the token, and a new token is issued after a lapse of a predetermined time interval. This interval is to prevent tokens from traveling too early when the text request frequency is low. This allows each communication control unit to allow data access (preparation for transmission) to each computer when circulating tokens. You can

The interval required to issue the token is the token issue management time (token issue cycle) Ts.

The token issuance management time Ts is the maximum token delay time in a large-scale system regardless of the scale of the communication system (the token maximum delay time is the time during which the token goes around the communication control device, regardless of which communication control device is used). It was generally fixed at a longer time, which is the time it takes if there is a text request. Token issue management time T
The reason for making s longer than the token maximum delay time will be described later.

In this way, the token issuance management time Ts is fixed uniformly according to the token maximum delay time in a large-scale system, in consideration of the scale of the communication system or the processing capacity of the computer. There is no consideration for rational setting, and efficient communication cannot be performed.

Therefore, conventionally, for example, JP-A-63-227149 has been used.
As in the loop communication system described in Japanese Patent Publication, there is proposed a means for providing a timer capable of arbitrarily variably setting the token issuance management time Ts according to the scale of the communication system and the processing capacity of the computer. It is evaluated that this means can contribute to the improvement of communication efficiency. However, when setting the token issuance management time Ts, it is unavoidable that the token issuance management time Ts is set longer than the maximum token delay time of each communication system, and further shortening of the token issuance management time Ts cannot be expected. In this conventional example, the reason why the token issuance management time Ts cannot be made shorter than the maximum token delay time will be explained in the section of the problem to be solved by the invention below.

[Problems to be Solved by the Invention]

FIG. 6 is a time chart showing an example of the operation of the loop communication system adopted in JP-A-63-227149, and FIG. 7 is a flowchart showing an example of token processing performed by the token issue management station. Is.

In FIG. 6, 2 _{1 to} 2 _n are communication control devices of respective computers, 2 ₁ is a communication control device of a main station that is the token issue management, and 2 ₂ to 2 _n are communication control devices other than the token issue management station ( Here it is assumed to be a slave station).

Further, R is reception, S is transmission, F ₁ is a token issued by the master station 2 ₁ , F _{2 to} F _n are slave stations 2 ₂ to 2 _n , and when there is a text transmission request, the token issuance name is the text. What was rewritten to the name of the slave station that requested transmission, "D ₂ →
₁ ”is the text to be transmitted from the slave station 2 ₂ to the master station 2 ₁ ,
“D _n → _n−1 ” indicates a text to be transmitted from the slave station 2 _n to the slave station 2 _n−1 .

This conventional example, the main station 2 ₁ performs the issue management of tokens. A program for reading and discarding tokens is provided regardless of whether it is a master station or a slave station, and the tokens are discarded when the token in the name of the station returns to the master station or the slave stations. In addition, there is so as to make a start-up Ts (token issuance management time) when the token also for whatever reason the name has come to the main station 2 _1.

An example of the operation will be described from the left side of FIG. 6, that is, Ts is timed up at the present time and the token F ₁ is generated in the main station 2 ₁ .

Token _{F 1} is received by the slave station _{2 2,} with the slave station _{2 2 "D 2}
When there is a request for text transmission of "→ ₁ ", the token issuance name is changed to F ₂ and the text "D ₂ → ₁ " is sent to the next station together with the token F ₂ . At this time, the token F ₂ follows the text transmission.

In this way, the token circulates through each station in sequence, and the text circulates with the token up to the destination station. If in this example, to reach the station of 2 _n, a text transmission request at each station of the communication control apparatus, and text transmission request in 2 _n stations (here to "D _n → _n-1") is Has
"D ₂ → 1" ... "D _n → _n-1 " and the text to be transmitted are concatenated one after another, followed by a token in the name of "F _n " and sent to the main station 2 ₁ station. To be At this time,
The maximum token delay time T _MAX is the time when the token makes one round with all the slave stations requesting text transmission.

In the main station 2 _1, (expressed this in reception indicated by the broken line block in the drawing) to take the text the own station is the destination address. At this time, the token returns to the main station 2 _1, so is other than own station name, discarding tokens is not performed, the text transmission request in to the remaining text tokens as it (the own station is here (This is because it exemplifies the case where there is no case), and the Ts activation is performed at this point by transmitting to the next station.

By repeating such a transmission operation, when the texts are successively received at the destination stations and no new text transmission request is made at each station, only the token remains (in FIG. 6, the text transmission request is finally made). Exemplifies the token “F _n ” in the name of the station which was made), and is discarded when it comes to the own station (here, 2 _n is an example). In this case, it does not come back a token to the main station 2 _1, after the token issuance management time Ts (Ts time up) the token is newly published as "F _1".

Then, after a token issued in the main station 2 _1, for example, even the token is sent to the slave station 2 ₂ to 2 _n in order, where if any text transmission request not, the main station 2 ₁ is a token issuing management source The token will be returned under the name of "F ₁ ", but at this time, the token will be discarded, Ts will be activated, and after Ts, a new token will be issued as "F ₁ " as in the above. The time required for the token to make a round at this time is the minimum time T _MIN .

Incidentally, the main station 2 _1, as long as the token comes repeatedly in another station name, it means that the text transmission is made, it is impossible to read and discard the token, for multiple issuing prevention of tokens in this case At the time of transmitting a token in the name of another station, Ts is restarted (in other words, Ts is restarted before Ts times out).

Of the above operation, a flow chart of FIG. 7 indicate the operation of the master station 2 _1. That is, the main station 2 _1, (when the token has been round the communication control device via the loop communication line) when the token received, the token is converted to issue the name of the other stations in the middle or issue the name of the own station performs a judgment of Taka, in the case of a self-station name performs a timer start-up of Shi discarded reading a token token issue management time Ts, to send a token of the main station 2 ₁ issue at the time is up. On the other hand, if it is in the name of another station, the timer for Ts is started as it is, and if there is no token received by the time up, the token is issued and transmitted as if the token was discarded during the process, and the token returned again before the time up. At this time (token reception), the token is transmitted again as it is, and the Ts timer is restarted (restarted).

In such a conventional technique, the token is the main station 2 ₁
Since it may disappear (in other words, it is discarded) by the time when comes, it is necessary to prepare for the token issuance at a certain point, and this is Ts activation. Then, conventionally, as a standard for starting Ts, it is set at the time when the token returns to the master station regardless of the names of the master station and the slave stations as described above. Further, the token issuance management time Ts must be set from the standpoint of preventing multiple issuance of tokens by taking into consideration whether the token is being discarded by a station on the way or returning to the main station.

Since the token maximum delay time T _MAX is the standard, at least Ts needs to be longer than T _MAX .

That is, it is necessary to secure Ts ≧ T _MAX + Ti (where Ti is a margin).

As described above, conventionally, the token issuance management time Ts is T
_I couldn't make it shorter than _MAX .

The present invention has been made in view of the above points, and an object thereof is a token ring loop communication system of this type,
The token issuance management time Ts is equal to the maximum token delay time T
It is intended to be much shorter than _MAX to further improve communication efficiency.

[Means for Solving the Problems]

According to the present invention, each of the communication control devices 2 ₁ to 2 _n of a plurality of computers to be subjected to data communication is connected as a station to each other via a loop communication line, and one of these communication control devices is a token. issue management station 2 _1, and the issued token is cyclically to each communication control device 2 ₂ to 2 n _... 2 ₁ order of connection from the token issuance management station 2 ₁ is put on the loop communication line, talk their communication When it comes to the control device, if there is a request for text transmission, the token issuance name is converted to the station name with the text transmission request, and after text transmission, the token is sent to the next communication control device and the text transmission without request, the loop communication system of the token ring system to send the token as it is in the subsequent communication control apparatus, communication control device of the token issuing management station 2 _1, A timer for setting a token issuing administrative time Ts required to be issued a token, if the token is returned to round in its own station issued name "F _1", the start the timer Shi read and discarded the token Token issue management time T
When s has elapsed, a new token is issued, and when the token is converted to the name issued by another station (any of F _{2 to} F _n ) and returns, regardless of whether there is a text transmission request at the own station. A token processing program for converting the token into a name "F ₁ " issued by the local station and loading it again on the loop communication line.

[Action]

 According to the above configuration, the following operation is performed.

An example of this operation will be described with reference to the time charts of FIGS. 3 and 4 used in the embodiment for easy understanding.

Figure 3 is an example of a case where text transmission frequency is low, the token issued by the communication control unit (master station) 2 ₁ as a token issuing management station, after transmitting the token directly with the own station issued name "F _1" A case where one communication control device makes a round is illustrated. Specifically, when there is no transmission request in the communication control device of any station in the process of the token circulating in the loop communication line, and when only the main station 2 ₁ requests a text (here, D ₁ → _n )
A case is illustrated.

In this case, start the timer and discard the token in the main station 2 ₁ (Ts starts) to the token issuance management time T
When s has elapsed, a token is newly issued under the name of its own station "F ₁ ", and this token is put on the loop communication line again and circulated between the communication control devices in order.

Figure 4 is a token issuing at the main station 2 _1, after transmission, the other station issues name while the token makes a round the communication controller (F ₂
~ Any of F _n ) and returns to the master station 2 ₁ (that is, a text transmission request is made by at least one of the communication control devices 2 ₂ to 2 _n , in this case, all the slave stations 2 _{1 2 to} 2 _n exemplifies the case of T _MAX in which there is a text transmission request), and the case where the main station 2 ₁ returns with its own station name “F ₁ ”.

In the case of the present invention, as shown in FIG.
_{If the} token name returned to ₁ is the name issued by another station, the token is converted to the name issued by the station “F ₁ ” (F in FIG. 4) regardless of whether there is a text transmission request at the station.
An example in which the name is converted from _n to F ₁ is shown), and it is put on the loop communication line again, and in this case, there is a large feature different from the conventional one in that it is not necessary to start the timer.

That is, in this way, when the token is recirculated based on the master station 2 ₁ , the token issuance name is always “F ₁ ” at _first , so that the slave stations 2 ₂ to 2 _n have their own token issuance name. Then the situation of returning to the own station (reception) disappears. As a result, the token is in the middle, in other words, slave station 2
_The situation of disappearing (reading off) in ₂ to _2n does not occur.

Therefore, it is not necessary to set the token issuance management time Ts in anticipation that the token will disappear on the way as in the conventional case, and the token issuance management time Ts can be set without being restricted by the maximum token delay time T _MAX. .

Then, the token issuance management time Ts can be set anew by reading out the token only when the token returns to the main station 2 ₁ only with the own station issuance name “F ₁ ”.
When the tokens are circulated, each communication control device may set the interval only by the minimum necessary margin Ti that gives a margin of data access (transmission preparation) to each computer (Ts = Ti). As a result, the token issue management time Ts can be made significantly shorter than the token maximum delay time T _MAX .

〔Example〕

An embodiment of the present invention will be described with reference to FIGS. 1 to 5 and 8.

FIG. 1 is a configuration diagram of the entire apparatus to which the present invention is applied, and includes a plurality of main computers (so-called host computers) 1 ₁ to
1 _n are connected to each other via the respective communication control devices 2 ₁ to 2 _n and the loop communication line 4. Each computer 1 ₁
˜1 _n and communication control devices 2 ₁ to 2 _n are connected via computer interface buses 3 _{1 to} 3 _n .

The communication control devices 2 _{1 to} 2 _n have general functions such as data transfer control with the computers 1 ₁ to 1 _n , frame monitoring on the communication line 4, text transmission / reception control, and a new token issue management function described later. .

Main station that manages token issuance (Token issuance management station)
In order to control the token issuance management, the communication control device as described above selects one of the communication control devices so that two or more communication control devices do not exist at the same time. The token issuing management station, there as a communication control device 2 _1, the communication control device 2 ₂ to 2 _n the other is a slave station that does not perform the token issuance management.

The communication control device 2 of the _main station, the token issuer, token discard, and a program for performing the token issuance management interval securing the token.

Further, all the communication control devices 2 ₁ to 2 _n are configured to be able to control text transmission and reception when a token comes to the own communication device (at the time of token reception). Then, in the case of text transmission, the token can be converted into the issuance name of the communication control device of the own station which has made the request and can be transmitted.

The transmitted text is captured by the reception when reaching the destination communication control device and removed from the loop. At this time,
When the communication control device at the receiving end does not make a transmission request, the token is transmitted as it is to the next communication control device.

Here, the token is a frame having a specific code pattern for giving a transmission right to each communication control device, and includes issued communication control device number information (issue name). The frame means a token and text.

Figure 2 is a block diagram of the communication control device 2 ₁ as a main station of the present embodiment. In Figure 2, computer interface bus 3 ₁ is connected to the computer-side control unit 8, looped communication line 4 is connected to the communication line control unit 9.

The computer-side control unit 8 and the communication line control unit 9 are connected to the arithmetic processing unit 5 via the internal interface bus 11.

In addition to the above, a program storage unit 6, a data storage unit 7, and a timer unit 10 are also connected to the arithmetic processing unit 5 via an internal interface bus 11.

It forms a token issuing management station serving the communication control device 2 ₁ of a hardware configuration of FIG. 2 as described above.

The communication control device 2 ₁ is a computer interface bus 3 ₁
Received by the computer-side control unit 8 to send commands and data transmitted from the computer 1 ₁ via a report to the processor 5 via the internal interface bus 11.

The arithmetic processing unit 5 uses a program in a program storage unit 6 in which a program for performing a predetermined transmission control process is stored in advance and a data storage unit 7 for temporarily storing transmission / reception data to interpret a transmission command, convert the format of transmission data, and the like. After the token is received by the communication line side control unit 9 or after the token issue management time Ts of the timer unit 10 is up, the transmission data is converted into a predetermined timing and format and transmitted to the communication line 4 as a transmission text. After that, the token issued by the local station is transmitted.

On the other hand, the text received from the communication line 4 is received by the communication line side control unit 9, converted into a predetermined timing and format, and reported to the arithmetic processing unit 5 via the internal interface bus 11.

The processor 5 in advance performs format conversion of the received data by the program and the data storage unit 7 performs a predetermined reception control process, the computer-side main computer via the computer interface bus 3 ₁ by the control unit 8 1
₁ , the reception report and the reception data are transmitted.

Here, the communication control device 2 of the _main station, is illustrated by the flow chart of the differences in the functions of the communication control device 2 ₂ to 2 _n of the slave Figure 5 and Figure 8.

The communication control device 2 of the _main station, the timer unit 10 to set the only station that issue management tokens, the token issuance management time Ts for the interval ensuring required for token issuance
And further includes a program as shown in FIG. That is, when the token comes after searching the communication control device _{2 1 ... 2 2 ~2 n ...} 2 1 via a loop communication line 4, and determines whether the token is self-station issues name,
In the case of the name issued by the own station, the token is read out, the timer is started, and when the token issue management time Ts has elapsed, the token is newly issued under the name issued by the own station. At this time, if there is a text transmission request at the local station, the text is transmitted and the token is subsequently transmitted. Also, when the token returns in the name issued by another station, regardless of whether there is a text transmission request in the own station, after converting the token to the name issued by the own station, it is put on the loop communication line again and the token is returned. To send. The above token processing program is stored in the storage unit 6.

On the other hand, the communication control device 2 ₂ to 2 _n of the slave station, the token issuer management function has no, as shown in FIG. 8, upon receiving the token from the previous station, if there is no text transmission request Is equipped with a program that sends the token as it is to the communication control device of the own station, converts the token into the name issued by the own station, and sends the token following the text transmission only when there is a text transmission request. The program is stored in its own storage unit.

Here, the operation of this embodiment will be described with reference to FIG. 3 and FIG.

The reference numerals used in the time charts of FIGS. 3 and 4 are defined in the same manner as in FIG. 6 described above.

As shown in FIG. 3, the frequency of occurrence of communication requests on the loop communication line is low, and the main station (communication control device of the token issuing management station) 2 ₁ and the slave station (communication control device other than the token issuing management station) 2 ₂ If there is no text transmission request in either or, if there is text transmission request only in the main station 2 _1, the main station 2 ₁ issued token F ₁ is the communication control device 2 ₂
After going through ~ 2 _n , it returns to the main station 2 ₁ , but at this time, the token is discarded. Then, when it is discarded, the timer is started (Ts is started), and when the token issuance management time Ts has elapsed, a new token is issued under the own station issuance name “F ₁ ”, and this is again sent to the loop communication line. The communication control devices of the respective stations are sequentially circulated between the stations.

Further, as shown in FIG. 4, tokens issued at the main station _{2 1,} after transmission, inter token communication control device ₍₂ 1 _{to 2} n)
If there is a text transmission request at a slave station (any of 2 ₂ to 2 _n ) during one cycle of the cycle, the main station 2 ₁ receives a token that is converted to another station issued name (F _{2 to} F _n ) on the way. Back,
In this case (here, the frequency of occurrence of communication requests is high, and text transmission requests from computers are sequentially generated in each slave station. As an example, all slave stations 2 ₂ to 2 _n have text transmission requests.
_In the case of _MAX, the main station 2 ₁ converts the token into its own station-issued name “F ₁ ” (in FIG. 4, the issue name is converted from F _n to F ₁ ). Example),
The token is not lost because it is put on the loop communication line again and the token is circulated again.

In the case of this embodiment, as shown in FIG. 4, when the issue name token back to the main station 2 ₁ is at another station name, with or without text transmission request in the local station 2 ₁ Part After converting the token into the name “F ₁ ” issued by the local station, the token is put on the loop communication line again, and in this case, there is no need to start the timer, which is a major feature different from the conventional one.

That is, in this way, when the token is re-circulated based on the master station 2 ₁ , the token issuance name is always “F ₁ ” at the beginning, so that the subordinate stations 2 ₂ to 2 _n can use the issue name under their own names. The situation of returning (receiving) to your own station disappears. As a result, the token is in the middle, in other words, the slave stations 2 ₁ to 2 _n.
It does not happen that it disappears (is discarded).

Therefore, it is not necessary to set the token issuance management time Ts in anticipation that the token will disappear halfway as in the conventional case, and the token issuance management time Ts does not have to be restricted to the maximum token delay time T _MAX .

Then, the token issuance management time Ts is limited to when the token issuance management time Ts is returned to the own station 2 ₁ only with the own station issue name “F ₁ ”.
Since “F ₁ ” can be read out and newly issued, each communication control device gives a margin of data access (transmission standard) to each computer when circulating the token, and the interval is set only by the minimum necessary margin Ti. It suffices (Ts = Ti). As a result, according to the present embodiment, the token issuance management time Ts is set to the token maximum delay time T _MAX.
Can be significantly shorter than.

That is, the token minimum delay time T shown in FIG. 3 and FIG.
_MIN and Token maximum delay time T _MAX may vary the processing capability of the scale and the computer system of the communication system, activation of the timer unit 10 of the token issuer management time Ts to carry out a token issuing management its name in the main station 2 ₁ issued The token issuance management can be performed without being affected by the scale of the communication system and the processing capability of each computer system, since the token F ₁ is received and discarded.

Therefore, according to the present embodiment, the token issuance management time can be significantly shortened compared to the conventional case, and
When such token issuance management time is repeatedly used, the total token issuance management time becomes much shorter than the conventional one. As a result, communication efficiency can be significantly improved.

[Effect of the Invention] According to the present invention, the token issue management time Ts of the loop communication system can be made significantly shorter than the token maximum delay time T _{MAX with a} simple configuration such as a timer and a token processing program. When the token issuance management time is repeatedly used, the total token issuance management time is much shorter than the conventional one, and as a result, the communication efficiency can be greatly improved. Moreover, regardless of the scale and processing capacity of the communication system, it is possible to contribute to improvement of communication efficiency of various small to large-scale communication systems.

[Brief description of drawings]

FIG. 1 is an overall configuration diagram of a loop communication system to which the present invention is applied, FIG. 2 is a block diagram of a communication control device for performing token issue management of the loop communication system according to an embodiment of the present invention, FIG. 3 and FIG. FIG. 4 is a time chart showing the operation of the loop communication system of the above embodiment, FIG. 5 is a flow chart showing the program contents of the communication control device of the token issue management station (main station) used in the above embodiment, and FIG. FIG. 7 is a time chart showing an operation example of a conventional loop communication system, FIG. 7 is a flow chart showing program contents of a communication control device which is a main station of the conventional loop communication system, and FIG. 8 is a token issue management station in the above embodiment. It is a flowchart which shows the program content of the communication control apparatuses other than. ₁ 1 ... 1 _n ...... _computer, 2 1 _{~2 n} ...... communication control device,
3 _{1 to} 3 _n ... Computer interface bus, 4 ... Loop communication line, 5 ... Arithmetic processing unit, 6 ... Program storage unit, 7 ... Data storage unit, 8 ... Computer side control unit,
9 ... Communication line side control unit, 10 ... Timer.

Claims (2)

[Claims] 1. A communication control device of a plurality of computers subject to data communication is used as a station to mutually connect via a loop communication line, and one of these communication control devices is used as a token issuing management station. , The issued token is put on the loop communication line and circulated to each communication control device from the token issuing management station in the connection order, and when the token comes around to its own communication control device, if there is a request for text transmission, the token Token is sent to the next communication control unit after the text transmission and the token is sent to the next communication control unit as it is if there is no request to send the text. In the ring loop communication system, the communication control device of the token issue management station is a token issue manager required to issue a token. When the token that has set the processing time Ts and the token has returned in the name issued by its own station, the token is discarded, the timer is activated, and a new token is issued when the token issuance management time Ts has elapsed. When issued and the token is converted to the name issued by another station and returned, the token is converted to the name issued by the local station regardless of whether there is a text transmission request at the local station, and then the loop communication line again. A loop communication system, comprising: 2. The loop communication system according to claim 1, wherein the token issue management time Ts set by the timer can be arbitrarily set.