JPS6113625B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an inter-computer transmission system for transmitting data between one parent computer and a plurality of slave computers.

This type of conventional computer-to-computer transmission method is the first
Data highway 11 on the ring as shown in the figure
A parent computer 12 and a plurality of child computers 13, 14, and 15 are connected to each other, and data transmission between the parent computer 12 and the child computers 13, 14, and 15 is performed through a common data highway 11. Alternatively, as shown in FIG.
4 and 15 are sequentially connected to lines 16, 17, and 1.
8, and data transmission is performed between the parent computer 12 and the child computer 1 through the line.
It is held between 3, 14, and 15.

In the conventional transmission systems shown in FIGS. 1 and 2, data is only transferred between a parent computer and one child computer at the same time on a series of transmission lines. Therefore, in order to perform data transmission between these computers with sufficient response time, data transmission must be sufficiently fast, and the transmission line 1
1, 16, 17, and 18 must be of high speed. As the distance between the parent computer and the child computers increases, the cost of the high-speed line and the cost of the interface with the line also increase. For this reason, there is a drawback that the overall cost is high in order to collect data from many child computers into one parent computer at a sufficient speed.

SUMMARY OF THE INVENTION An object of the present invention is to provide an inter-computer transmission system that can be constructed at a relatively low cost and that enables data transfer between a parent computer and a plurality of child computers at high speed.

According to this invention, a relay computer is provided between a parent computer and a plurality of child computers,
The relay computer and the parent computer are connected through a high-speed line, and the relay computer and each child computer are connected through separate low-speed lines. The relay computer is equipped with a buffer corresponding to each child computer to store the child computer's data, especially the latest data.When a parent computer requests data from a child computer, the relay computer responds to the data request from the parent computer. retrieves data from the buffer of the corresponding child computer and transfers it to the parent computer. If a relay computer makes one data request to a child computer, it retrieves data from the corresponding child computer and stores it in a buffer even if there is no subsequent request to that child computer. Therefore, when the parent computer next requests data from its child computer, the relay computer reads out the buffer for the child computer and transfers it to the parent computer. After a parent computer requests data from one child computer, it immediately requests data from other child computers to collect a large amount of data. During this time, the relay computer collects data in parallel to each child computer through separate low-speed lines, asynchronously with the data transfer to the parent computer. If the parent computer does not request data transmission for a predetermined period of time, the relay computer stops requesting data from its corresponding child computer. In this way, the overall configuration can be made inexpensive, and data transmission between a large number of child computers and the parent computer can be performed at high speed as a whole.

For example, as shown in FIG.
and child computer 1 set up apart from this.
Relay computer 21 between 3, 14, and 15
will be provided. The relay computer 21 and the parent computer 12 are connected by one high-speed line 22, and the child computers 13, 14, and 15 are connected by low-speed lines 23 and 2, which are slower than the high-speed line 22, respectively.
4 and 25 are connected separately. The relay computer 21 includes child computers 13, 14, 15.
Buffers 26, 27, 28 are provided correspondingly to the counters 31, 32, 3 for detecting that there is no request for a predetermined period of time.
3 is provided.

Data transmission between the parent computer 12, the relay computer 21, and the child computers 13, 14, and 15 operates as follows. When the parent computer 12 issues a data request to the child computer 13 at a certain time _t1 as shown in FIG. A request for data α containing the contents is issued as shown by the solid line in FIG. Transfer this to child computer 13
receives the data α and relays the data α to the relay computer 21.
The relay computer 21 transfers the data α to the parent computer 12 at time _t3 after receiving it. Parent computer 12 receives this at time _t4 .

After that, the relay computer 21 automatically issues a request for the data α to the child computer 13 from time t ₃ even if there is no request for data α from the child computer 13 from the parent computer 12, and the child computer 13 automatically requests the data α from time t _5. The next data α is sent to the relay computer 21, and the relay computer 21 stores the data in the corresponding buffer 26 at time _t6 . Data requests from the relay computer 21 for data α of the child computer 13 after that point _t3 are sent to the parent computer 12.
This is done asynchronously.

Next, when the parent computer 12 sends a request for data α from the child computer 13 to the relay computer 21 at time t ₇ , the latest data is read from the buffer 26 and sent back to the parent computer 12 at time t ₈ . It's coming. The relay computer 21 receives a request for data α from the child computer 13 from the parent computer 12 at time _t9.
If the parent computer 12 does not give a request for the data α of the child computer 13 for a certain period of time T ₁ in response to the specified period of time T ₁
The counter 31 detects that the period has elapsed. That is, each time the parent computer 12 makes a request for the data α of the child computer 13, the counter 31 is reset and a new counting of clocks of a certain period is started, until the counter 31 reaches the count value corresponding to time _T1 . Parent computer 12
Even if there is no request for data α from the child computer 13, the relay computer 21 sends the data α to the child computer 13 as shown at time _t11.
make a request. However, when a period of time T ₁ has elapsed since the last request for data α from the parent computer 12, the relay computer 21 automatically stops sending requests for data α from the child computer 13 from that time point t ₁₀ onwards. do.

Since the transmission line 22 between the parent computer 12 and the relay computer 21 is connected by a high-speed line, the relay computer 21
, send a data request to buffer memory 2.
Reading from 6, 27, and 28 can be done at extremely high speed and in a short time. Therefore, after issuing a data request to a child computer, a series of data (obtained by multiple requests) for that child computer can be read. In the process of obtaining data, it issues data requests to other computers. For example, in FIG. 4, when a data request is issued to the child computer 14 at time t _{12 after time t 1 and} the latest data of the child computer 14 is received from the buffer 27 of the relay computer 21, the child computer 14 is sent to the child computer 14 at time t ₁₃ . Issue a data request for 15. When the latest data of the child computer 15 is received from the buffer 28 of the relay computer 21, a data request is issued to the child computer 13 at time _t14 . In this way, data requests are sequentially issued to the child computers 13, 14, and 15 at high speed. On the other hand, the relay computer 21 is connected to the child computers 13, 14, 1
As mentioned earlier, when there is a data request for 5, the data request is automatically sent to the corresponding child computers 13, 14, 15 using the solid line,
Collect data from the child computer as shown by the dotted and broken lines, respectively, and perform buffer 2.
6, 27, and 28 respectively. Further, each time a request is made to each child computer by the parent computer 12, the corresponding counters 31, 32, and 33 are reset and new counting is started. While the parent computer 12 is sequentially sending out data requests for each child computer to the relay computer 21, the data of each child computer is sent to the relay computer 21.
1, data is fetched in parallel, and the fetch can be performed until the parent computer 12 makes the next data request to the same child computer. Therefore, the parent computer 12 can acquire data from all child computers at a high speed. Moreover, it is only necessary to connect the parent computer 12 and the relay computer 21 with a high-speed line, and the relay computer 21 and the child computers 13, 14,
15 may be of low speed and can be constructed at low cost. Furthermore, if the data request from the parent computer stops without continuing for a certain period of time, the relay computer 21 stops requesting data from the corresponding child computer, so the type of data requested by the parent computer from the corresponding child computer is When the transmission load changes, it is possible to prevent the transmission load on the low-speed line between the child computer and the relay computer from increasing due to unnecessary transmission. Requests from the parent computer to the relay computer may or may not be made at regular intervals, and as shown in this example, data requests from the parent computer to all child computers may be made at once. Although data for one child computer is collected on the relay computer only once each time, it is not necessary to do so. In other words, the relay computer may request the child computer more quickly than in the case of FIG. 4.

Further, when a request from a parent computer to a child computer is stopped, the fixed time period _T1 for stopping a data request from a relay computer to the corresponding child computer can be freely selected. When connecting a relay computer and a slave computer via a low-speed line, not only one slave computer but also a plurality of slave computers may be connected to one low-speed line in an interlocking manner as described with reference to FIG. In any case, a plurality of low-speed lines are connected to the relay computer so that parallel processing can be performed.

As described above, according to the computer-to-computer transmission system according to the present invention, by using one relay computer, if there is one data request to the child computer, the data request from the relay computer to the child computer is By automatically issuing data, in other words, by making data requests independently of the parent computer, the relay computer can take in data from multiple child computers in parallel and create a buffer of the latest data. can be retained. In this case, the transmission line between the relay computer and the child computer can be a low-speed one and can be constructed at low cost, while only one high-speed line can be provided between the parent computer and the relay computer. Bye. In addition, when there are no data requests from the parent computer to the relay computer for a certain period of time, data import requests to the child computer are automatically stopped, allowing for changes in the type of data requested from the parent computer to the child computer. This can prevent the transmission load between the child computer and the relay computer from increasing.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing a conventional computer-to-computer transmission system, FIG. 2 is a block diagram showing another example of the conventional computer-to-computer transmission system, and FIG. 3 is a block diagram showing an example of the computer-to-computer transmission system according to the present invention. 4 are diagrams showing the flow of data transmission to explain its operation. 12... Parent computer, 13, 14, 15... Child computer, 21... Relay computer, 22
...High speed line, 23, 24, 25...Low speed line,
26, 27, 28... Batsuhua, 31, 32, 33
...Counter for time counting.

Claims (1)

[Claims] 1. In an inter-computer transmission system for transmitting data between one parent computer and a plurality of child computers, a relay computer is provided between the parent computer and the plurality of child computers; The relay computer and the parent computer are connected by a high-speed line, and the relay computer and each of the child computers are connected by a low-speed line whose transmission speed is lower than the high-speed line, and the relay computer A buffer corresponding to each of the above-mentioned slave stations and a buffer are provided for each of the above-mentioned child computers, and each time a data request from the above-mentioned parent computer to a child computer is made, the child computer corresponding to that child computer starts counting anew the elapsed time. A counting means, and in response to a data request from the child computer specified by the parent computer, extracts data from the child computer and transfers it to the parent computer, and once data is requested from the parent computer, the parent computer does not rush after that. Even if the data request is automatically issued to the corresponding child computer at least once, the data is retrieved and stored in the corresponding one of the buffers mentioned above, and in that state, the corresponding child computer When a data request for a buffer arrives from the parent computer, the corresponding data in the buffer is read and transferred to the parent computer, and when the counting means counts a predetermined elapsed time, automatic transmission of data requests to the corresponding child computer is stopped. and an inter-computer transmission method in which the parent computer requests data from other child computers during a plurality of data requests for a required series of data from the one child computer.