JPH11249976A - Data transfer control system in client-server system - Google Patents

Abstract

(57) [Summary] [PROBLEMS] In data transfer on a client-server system, avoid duplicate transfer of the same data to reduce the communication load, waste communication capacity due to abnormal data transfer, and perform abnormal data transfer on a network. To prevent propagation at SOLUTION: In a client-server system having at least one server, a plurality of clients, and a network for connecting them and exchanging data, each client intervenes between the client and the server on the network. A request for data transfer from the server is received in place of the requested server, a request for transfer of the same data from the plurality of clients to the same server is sent together to the server, and response data from the server for the transfer request is received. And an exchange for distributing the response data to each client.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a data transfer control system in a client server system having a plurality of servers, a plurality of clients, and a network for exchanging data by connecting the plurality of servers and the clients. This is a controller that performs loop control of the above, and is suitably used for a plant control monitoring system in which a client is an operator station.

[0002]

2. Description of the Related Art FIG. 7 shows a document: Chaia Shen.
and Ichiro Mizunuma. “RT-C
RM: Real-Time Channel-base
d Reflective Memory. "IEEE
RTAS 1997. 1 is a diagram showing a configuration of a large-scale client-server system using the technology described in FIG. The network Na is connected to m client nodes C1, C2,..., Cm. The network Nb includes n server nodes S1, 1, S1, 2,.
, S1, n, and n server nodes S2, 1S2, 2,..., S2, n are connected to the network Nc. Network Na and Nb, Na and Nc are gateways (or routers, bridges, etc.)
Are connected to each other. Network Nb and N
c may be different in the number of server nodes.

Each server node has a memory area and middleware. The contents of the memory area can be rewritten from time to time by, for example, certain hardware or software in the server node. The middleware in the server node receives a request from another node and returns a response to the request. Here, the request means that the data in the memory area of each server is
oS class (quality of service,
(Communication quality). FIG. 8 shows the structure of the request message.

[0004] Here, the response is to transfer each data to the node that issued the request in accordance with the specified QoS class in accordance with the request. Qo
The S class is, for example, a set including the following parameters. -Data transfer cycle / or minimum transfer interval-Data arrival delay-Flag indicating whether transfer is periodic or update-based Each client node has an application and middleware. The application requests the middleware to transfer data in the memory area of a certain server to its own node. The middleware of the client node transfers this request to the corresponding server as a request message shown in FIG. With this request, the response message sent back from the server to the client node is passed to the application via the middleware. FIG. 9 shows the structure of the response message.

[0005] Each gateway passes request and response messages, which are transferred across the networks to which it connects, between networks. Although an example including three networks has been described here, a system in which four or more networks are generally connected by a gateway or the like may be used.

[0006]

Since the conventional client server is configured as described above, requests for transferring the same data from a plurality of client nodes to a certain server node are generated at the same time. If this server node receives each request message separately,
Returns a response for each client node separately. As a result, the same response processing is performed a plurality of times, and there is a problem 1 that the load for the response of the server node becomes higher than an actually necessary amount.

There is also a problem 2 that the transfer load of the network to which the server node is connected becomes higher than an actually necessary amount due to the same cause as the problem 1.

[0008] Further, there is a problem 3 that, due to the same cause as the problem 1, the load of the server node for receiving the request message becomes higher than an actually necessary amount.

Also, when a certain server node transfers data that is not actually requested to a certain client node due to a failure or the like, or when a request is made but the requested period is shorter than the requested period. In the case where data is transferred periodically, there is a problem 4 that the process of receiving a response from the corresponding client node becomes higher than expected by the client, which may hinder correct operation of the client node.

[0010] Further, for the same reason as in the problem 4, the transfer load of the network to which the corresponding client node is connected becomes higher than expected by each node in the corresponding network, and prevents the proper operation of the corresponding node. There is a problem 5 that there is a fear.

SUMMARY OF THE INVENTION The present invention has been made to solve the above-mentioned problems, and in the data transfer on a client-server system, avoids duplicate transfer of the same data, thereby reducing the communication load and preventing abnormal data transfer. An object of the present invention is to provide a data transfer control system in a client-server system that can prevent waste of communication capacity and propagation of abnormal data transfer on a network.

[0012]

According to a first aspect of the present invention, there is provided a data transfer control system in a client-server system comprising at least one server, a plurality of clients, and a network for connecting them and exchanging data. A client-server system having a client-server system that intervenes between the client and the server on the network, receives a data transfer request from each client in place of a requested server, and transfers the same data from a plurality of clients to the same server. The server is provided with an exchanger that collectively transmits requests to the server, receives and holds response data from the server in response to the transfer request, and distributes the response data to each client.

Further, the data transfer control system of the client server system according to the second configuration of the present invention, when combining the requests from the plurality of clients into one,
The communication quality required for the server to be received, the communication quality that satisfies all of the communication quality requested by the plurality of clients,
The received response data from the server is distributed to each client at the communication quality requested by each client.

[0014]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Embodiment 1 FIG. 1 is a diagram showing an internal configuration of the exchanger according to the first embodiment of the present invention. It is assumed that the arrangement of the exchanger in the system is the same as the arrangement of the gateway in the conventional technology (FIG. 7). A request message sent from a certain client node to a certain server node is received by a request receiving process 1 of an exchanger arranged between the server and the client.

FIG. 2 is a diagram showing a data structure of the request table according to the first embodiment of the present invention. Each table element in this table holds a copy of the latest request from one client to one server. The table elements in the table form a one-way list among all the table elements corresponding to the same client using the next_c pointer. The table element at the head of the list is indicated by each pointer of the client list indexed by the client identifier. The next_c pointer of the table element at the end of the list indicates a NULL value. N of other table elements
The ext_c pointer points to the next table element in the list. As a result, all table elements corresponding to a specific client, that is, all requests from this client can be sequentially accessed for each server.

Similarly, the table element in the table is n
A one-way list is formed between all table elements corresponding to the same server using the ext_s pointer. The table element at the head of the list is indicated by each pointer of the server list indexed by the server identifier. The next_s pointer of the table element at the end of the list indicates a NULL value. Ne for other table elements
The xt_s pointer points to the next table element in the list. As a result, all table elements corresponding to a specific server, that is, all requests to this server can be sequentially accessed for each client.

The request message received by the request receiving process 1 is passed to the request merging process 2. The request merge process 2 holds this request message as the latest copy of the request from the client to the server. That is, the content of the corresponding table element in the request table 4 is rewritten so as to be equal to the content of the received request. The previous value in the table element is initialized to a certain value that the data cannot take. For other previous values,
A description will be given of the response transmission process.

Next, request merge processing 2 performs request merge processing 2 for requests from all clients to the server. FIG. 3 is a diagram showing an algorithm of the request merge process. FIG. 4 is a diagram showing an example in which the request merge processing is performed according to the algorithm of FIG. In addition, each pointer and the number of data are omitted. The QoS class is described separately for each parameter (period / minimum interval, arrival delay, flag). The unit of time is not particularly defined here.

In the figure, a), b) and c) indicate requests from clients C1, C2 and C3, respectively. The shaded record indicates that there is a record whose address matches the record being requested from another client.

A request for the address 0100 has been issued from each of the clients C1 and C2. Both are periodic transfers, and the periods are 100 and 200, respectively.
Since the arrival delays are 50 and 100, respectively, the parameters satisfying both at the same time are a flag with periodic transfer, a period of 100, and an arrival delay of 50. The parameters of the QoS class for the address 0100 of the transmission data list shown in d) are set as described above.

A request for the address 0120 has been issued from each of the clients C1 and C3. The QoS class parameter of the request from C1 has a period of 50
0, arrival delay is 250, and flag is periodic transfer. On the other hand, those from C3 have a minimum interval of 200, an arrival delay of 100, and a flag update transfer. To satisfy both at the same time, each parameter in the transmission data list must be
Regarding the cycle / minimum interval and the arrival delay, 200 and 100 of C3 are respectively taken. However, for the flag, since C1 is the cyclic transfer, the cyclic transfer is similarly performed.

A request for the address 0130 has been issued from each of the clients C2 and C3. In this case, since both are updated, the flag of the transmission data list is also updated.

As described above, the transmission data list generated by the request merge processing 2 is transmitted to the corresponding server as a request message. At this time, the identifier of the corresponding exchanger is stored in the client identifier of the request message.

Next, the operation of the exchanger in response to a response message transmitted from the server node will be described. Response messages from each server are received by a response receiving process 5. The response receiving process 5 writes each data in the response message on the intermediate memory area 6.

FIG. 5 is a diagram showing a data structure of an intermediate memory area for a certain server. Intermediate memory area 6
Is prepared for every server connected to the exchanger. The number of data (num) in the intermediate memory area 6 is the number of data in the latest response message arriving from the server. The address is an address in the server of each data in the response message.

The response transmission processing 7 refers to the data in the intermediate memory area 6 and the contents of the request table 4 respectively,
Generate a response message to each client and send it to the client. Response transmission processing 7
This is a parallel process consisting of a plurality of tasks prepared for each class. If a QoS class that does not exist before is generated by the request merging process 2, a task corresponding to the QoS class is generated in the response transmission process 7. In addition, when the request for the QoS class that has been present no longer exists by the request merge process 2, the task in the response transmission process 7 corresponding to the QoS class is stopped or deleted. Each of these tasks is executed with an execution priority according to the corresponding QoS class. In the present invention, the relationship between each parameter of the QoS class and the task priority is not particularly defined.

FIG. 6 is a diagram showing an algorithm of a task in a response transmission process prepared for each QoS class. Each task is activated at a cycle equal to the cycle / minimum interval of the QoS class, and performs the processing of FIG. With this process,
If the flag of the relevant QoS class is periodic transfer,
For each set of all client servers,
From the client to the server, all of the data requested in the QoS class that is satisfied by the QoS class and arriving at the intermediate memory area 6 are added to the response message, and the response message is sent to the client. Send. If the flag of the relevant QoS class is the transfer at the time of updating, for each of the set of servers of all the clients, the corresponding client to the relevant server, and the Q satisfied by the relevant QoS class
Of the data requested in the oS class, all the data arriving at the intermediate memory area 6 and not matching the corresponding previous value in the request table are added to the response message, and this response message is added. Send to the client. At this time, the previous value is rewritten to the value of the data.

As described above, according to the present invention, requests for the same data from the plurality of client nodes to the same server are combined into one request by the exchanger. Therefore, it is possible to reduce the transmission load and the reception load caused by redundantly transferring the same data in the server node, and it is also possible to minimize the transfer load of the network.

In the merging process, the communication quality required of the server is a communication quality that satisfies all the requests of a plurality of clients requesting the same data. Since the distribution is performed according to the communication quality requested by the client, the communication load of the entire network can be reduced while satisfying the communication quality requirement of each client.

Further, according to the present invention, when a certain server node attempts to transfer data that is not actually requested to a certain client node, or when a requested time is shorter than the requested time. Even when the data is to be transferred periodically, the exchanger transfers the requested data to each client only at the requested cycle. Therefore, it is possible to prevent abnormal data transfer on the network, especially waste of communication capacity due to excessive data transfer, and propagation of abnormal data transfer on the network.

In the present embodiment, the case where the present invention is applied to one-way data transfer from a server to a client has been described. However, it is assumed that the exchanger is bidirectional and is applied to bidirectional data transfer. Can also.

[0032]

According to the data transfer control system in the client-server system according to the first configuration of the present invention, a client having at least one server, a plurality of clients, and a network for connecting them and exchanging data. In the server system, the server intervenes between the client and the server on the network, receives a data transfer request from each client in place of the requested server, and receives a transfer request of the same data from the plurality of clients to the same server. It is provided with an exchanger that collectively transmits the data to the server, receives and holds response data from the server in response to the transfer request, and distributes the response data to each client. Avoid duplicate data transfer of the same data Is, it is possible reduce the data transfer load, abnormal data transfer, especially waste or communication capacity due to excessive data transfer, it is possible to prevent transmission on the network of abnormal data transfer.

According to the data transfer control system in the client-server system according to the second configuration of the present invention, when the requests from the plurality of clients are combined into one, the communication quality required for the server to be transmitted is determined by the plurality of clients. Communication quality that satisfies all of the communication qualities requested by the client, and the response data received from the server is distributed to each client at the communication quality requested by each client. This can be sufficient and sufficient for each client, and can reduce the data transfer load of the entire network while satisfying the requirements of each client.

[Brief description of the drawings]

FIG. 1 is a diagram showing an internal configuration of an exchanger according to a first embodiment of the present invention.

FIG. 2 is a diagram showing a data structure of a request table according to the first embodiment of the present invention.

FIG. 3 is a diagram illustrating an algorithm of a request merge process.

FIG. 4 is a diagram illustrating an example in which a request merge process is performed according to the algorithm of FIG. 3;

FIG. 5 is a diagram showing a data structure of an intermediate memory area for a certain server.

FIG. 6 is a diagram illustrating an algorithm of a task in a response transmission process prepared for each QoS class.

FIG. 7 is a diagram showing a configuration of a conventional client server system.

FIG. 8 is a diagram showing a configuration of a request message.

FIG. 9 is a diagram showing a configuration of a response message.

[Explanation of symbols]

1 request reception processing, 2 request merge processing, 3 request transmission processing, 4 request table, 5 response reception processing, 6 intermediate memory area, 7 response transmission processing, 10 exchangers.

Claims (2)

[Claims] 1. A client-server system having at least one server, a plurality of clients, and a network for connecting them and exchanging data, interposed between the clients and the server on the network, A data transfer request from a client is received in place of a requested server, transfer requests for the same data from a plurality of clients to the same server are sent together to a server, and response data from the server to the transfer request is sent to the server. A data transfer control system for a client-server system including an exchanger that receives and holds the response data and distributes the response data to each client. 2. Requests from the plurality of clients are
At the time of putting together, the communication quality required for the server to be received is set to the communication quality that satisfies all of the communication qualities requested by the plurality of clients, and the received response data from the server is sent to each client. 2. The data transfer control system for a client server system according to claim 1, wherein the data is distributed with the requested communication quality.