JP2000276365A - Client server system and monitoring system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To obtain a client server system(CS) for executing communication by connecting plural servers to plural clients which can recognize only a single IP address for these servers by using a TCP/IP protocol and a monitoring system to be applied to the CS. SOLUTION: The same MAC address(MAC-A) is applied to plural servers 1, an independent MAC-A is applied to each of plural clients 2, and at the time of initializing when one of plural servers 1, 1a starts communication, whether the same address as the MAC-A of the server itself flows on a bus or not is judged. When the same address as the MAC-A of the server itself does not flow on the bus, the server completes the initialization, starts communication with the client, and when the same address as its own MAC-A flows on the bus, stops the initialization.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a client-server system and a monitoring system, and more particularly to an independent MAC address in which a plurality of servers can recognize only a single MAC address and an IP address for the plurality of servers. The present invention relates to a client-server system for performing communication by connecting one client among a plurality of clients having an IP address and a TCP / IP protocol, and a monitoring system to which the client-server system is applied.

Recently, there has been a remarkable spread of computer networks typified by the Internet. Among them, a computer called a server that controls the functions of the entire computer network and its own function using the functions of the server The client-server system, which is a hierarchical computer network composed of computers called clients that realize the functions, has become mainstream, and has been applied to extremely wide fields such as office networks and monitoring system networks. ing.

[0003] In a client-server system, a communication protocol called TCP / IP protocol is standard.

[0004] Here, TCP stands for Transmission Control P.
OSI (OpenSystem Interconnectio)
n: Open system interconnection) A protocol corresponding to the transport layer which is the fourth layer of the reference model. IP is In.
Abbreviation of ternet Protocol, a protocol corresponding to the network layer which is the third layer of the OSI reference model.

[0005] In a client / server system conforming to the TCP / IP protocol, an end server is used.
An IP address is assigned so that the system can be uniquely identified, and a MAC address (MAC is Media Access Control) so that the end system can be uniquely identified physically.
trol stands for: media access control. ) Is allocated.

[0006] Therefore, in order to communicate with a server having a duplex configuration, it is generally necessary for a client to be able to recognize two individual IP addresses and MAC addresses corresponding to each server.

FIG. 10 shows a basic configuration of a client server system in which a server is duplicated.

In FIG. 10, reference numerals 1 and 1a denote redundant servers, in which the server 1 is called a 0-system server in the sense of an active server, and the server 1a is 1 in a sense of a standby server.
It is sometimes called a system server. That is, since the server 1 and the server of the 0 system indicate the same thing, and the server 1a and the server of the 1 system indicate the same thing, in the following description, "server 1", "server 1a" and "0 system" Server "and" first system server ".

Reference numeral 2 denotes a client.

Reference numerals 3 and 3a denote buses. The bus 3 is a bus through which the client 2 communicates with the server 1, and the bus 3a is a bus through which the client 2 communicates with the server 1a.

Reference numeral 4 denotes a path through which the server 1 accesses the client 2 via the bus 3;
a is a path through which the client 2 accesses the server 2 through the bus 3a, 5 is a path through which the client 2 accesses the server 1 through the bus 3, and 5a is a path through which the client 2 accesses the server 2 through the bus 3a. 1a.

FIG. 11 is a diagram showing communication between the client and server in the configuration of FIG.

Since the components in FIG. 11 are exactly the same as those in FIG. 10, the description of the reference numerals is omitted.

In the client-server system shown in FIG. 11, the client 2 can recognize two individual IP addresses and MAC addresses corresponding to two servers that are communication destinations.

First, when a message is transmitted from the client 2 to the server 1, the client 2 and the server 1
After the communication is confirmed with the server 1, a unique MAC address is set, and a message composed of a packet called a MAC frame is transmitted to the server 1. In the data part of this I
The data is converted into an IP datagram with a P address added thereto and distributed to the server 1. Similarly, when transmitting a message from the client 2 to the server 1a, the unique M
An AC address is set, a message composed of a packet called a MAC frame is transmitted to the server 1a, and an IP address or the like is added to a data portion of the message, converted into an IP datagram, and distributed to the server 1 .

Therefore, it is possible to communicate only with the server 1 via the path 5 or to communicate only with the server 1a via the path 5a, or to communicate with the server 1 via the path 5 and the path 5a. 1 and the server 1a at the same time. The upstream and downstream communication paths forming this communication are shown by solid arrows and broken arrows in FIG.

Particularly, when communicating with the server 1 and the server 1a simultaneously via the path 5 and the path 5a,
Since the same information held by the client 2 can be simultaneously transmitted to the two servers, the reliability of the client's information management on the server side is increased.

However, in order to realize the above function, two paths, a bus 3 for the client 2 to access the server 1 and a bus 3a for the client 2 to access the server 1a, It is necessary, and at the same time, it is necessary to provide the client 2 with two sets of transmission / reception circuits in order to communicate with the two servers.

Therefore, the configuration shown in FIG. 10 has a disadvantage that the system scale becomes large.

For this reason, there has been a movement to configure a client-server system including a single server and a single client in response to a demand to reduce the system scale.

However, from the reflection that it is difficult to keep the reliability of a practical system high because a single server is used, a duplicated server and two individual MAC addresses and IP addresses corresponding to each server are assigned. While maintaining the same functions and reliability as a client-server system composed of clients that have the same client, a server with a duplex configuration and a client that can recognize only a single MAC address and IP address for two servers The realization of a configured client-server system is strongly desired.

That is, as a client / server system of server duplication and single client which is a target of the present invention, in a system form as shown in FIG.
Client / server that can only recognize addresses
The realization of a system is desired.

[0023]

2. Description of the Related Art In FIG. 1, reference numerals 1 and 1a denote redundant servers.
The server 1a may be referred to as a first-system server.

Reference numeral 2 denotes a client.

Reference numeral 3 denotes a bus, and the client 2 communicates with the server 1 and the server 1a via the bus 3.

Reference numeral 4 denotes a path through which the server 1 accesses the client 2 via the bus 3;
a is a path for accessing the client 2 via the bus 3, and 5 is a path for the client 2 to access the bus 3 such that the client 2 accesses the server 1 and the server 1a via the path 5. I want to make it.

Conventionally, in the configuration of FIG. 1, a global MAC address and a global I
Assuming that a P address is assigned, each server or client does not check the MAC address flowing on the bus at the time of initializing the start of communication.

[0028]

Therefore, if it is attempted to connect a duplicated server to a client that can recognize only one IP address of the server on the premise of the prior art, for example, the server 1 and the client 2 will be connected. When the server 1a tries to communicate with the client 2 during communication, the server 1a apparently can also be initialized, so that the server 1a also tries to start communication with the client 2 as shown in FIG. In the diagram (part 1) for explaining the problem of the prior art in the configuration, a communication error occurs as indicated by a thick broken cross mark X, and the communication between the server 1 and the client 2 also occurs between the server 1a and the server 1a. Communication between the clients 2 is also disabled.

FIG. 13 is a diagram (part 2) for explaining the problems of the prior art in the configuration of FIG. 1 and clarifies the operation of each of the duplicated servers in the prior art.
2 will be described in some detail.

In FIG. 13, first, the system 0 server starts communication. That is, the initialization process is performed, and when the initialization is successful, its own IP address is assigned and the TCP / I
A message conforming to the P protocol is sent to the bus 3. At this time, since the system 1 server is not communicating, the communication between the system 0 server and the client can be performed.

It is assumed that the server of system 1 attempts to initialize while the server of system 0 communicates with the client. At this time, even if the 0-system server is communicating with the client, the 1-system server also succeeds in initialization based on the conventional technology. The first-system server that has been successfully initialized assigns its own IP address and sends out a message conforming to the TCP / IP protocol onto the bus.

At this time, telegrams of the servers of both systems collide on the bus, and all of the server of the system 0, the server of the system 1 and the client 2 cannot communicate.

[0033] In view of the above problems, the present invention provides a method in which a plurality of servers and one of a plurality of independent clients that can recognize only a single MAC address and an IP address with respect to the plurality of servers are connected to a TCP / IP protocol. It is an object of the present invention to provide a client-server system that performs communication in conformity with the standard and a monitoring system to which the client-server system is applied.

[0034]

SUMMARY OF THE INVENTION As will be described later, the technique of the present invention provides a client server having a dual server.
Although the present invention is not limited to the server system, for the sake of simplicity, the description will be given with an example of a client server system in which the server is duplicated.

According to the present invention, 0 is added to the duplicated server.
In both the system and the system, the same MAC address and the same IP address are assigned, and the server that intends to communicate searches the bus at the time of initialization processing to determine whether the same MAC address as its own MAC address is flowing on the bus. When the same MAC address as the own MAC address is not detected, the initialization is regarded as successful, the LAN card is activated, communication with the client is started, and when the same MAC address as the own MAC address is detected. Assuming that the initialization has failed, it does not activate its own LAN card.
This is a technique for activating its own LAN card for the first time when the same MAC address as the address is no longer detected.

According to the present invention, the duplicated server searches whether or not the same MAC address as its own MAC address flows on the bus, and does not detect the same MAC address as its own MAC address. Only activates its own LAN card when it detects the same MAC address as its own MAC address.
Since only one server is given the authority to connect to the bus by not activating the card, even if the same MAC address is assigned to the duplicated server, both the duplicated servers communicate simultaneously. Never. Therefore, only a single MAC address can be recognized for a duplicated server, and a client having only a single port can communicate with any server on a single bus.

As a result, the size of a duplicate server and a single client / server system can be reduced, and the reliability of the client / server system can be kept high because the server is duplicated. it can.

[0038]

DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 shows a dual server, single client client server,
System.

The configuration of FIG. 1 has been described in the section of the prior art because of the necessity in the flow of the description, and therefore, the description of the configuration and reference numerals will be omitted here to avoid duplication.

FIG. 2 is a diagram for explaining means for making only one server communicable in the configuration of FIG.

In FIG. 2, reference numerals 1 and 1a denote redundant servers. Also in this case, the server 1 may be referred to as a system 0 server, and the server 1a may be referred to as a system 1 server.

Reference numeral 2 denotes a client.

Reference numeral 3 denotes a bus.

Reference numeral 4 denotes a path through which the server 1 accesses the client 2 via the bus 3;
a is a path for accessing the client 2 via the bus 3, and 5 is a path for the client 2 to access the bus 3.

Reference numeral 11 denotes a communication control unit provided in the server 1, and reference numeral 11a denotes a communication control unit provided in the server 1a.

The same MAC address and IP address are assigned to the server 1 and the server 1a.

Here, the description will be made on the assumption that the servers of both systems are not communicating with each other at first, and the server of system 0 tries to start communication first.

In the server 0, the communication control unit 11 starts the initialization process and searches for a MAC address flowing on the bus 3.

At this time, since the server of the system 1 is not communicating, the server of the system 0 does not detect the same MAC address as its own MAC address.
Therefore, the server of system 0 determines that communication is possible, and
After setting its own MAC address and IP address and confirming communication with the client 2, communication with the client 2 is started.

While the 0-system server is communicating with the client 2, the 1-system server attempts to initialize. At this time, the server of the first system also searches for the MAC address flowing on the bus 3 in the initialization processing in the communication control unit 11a. At this time, since the server of system 0 is already communicating with the client 2, the same MAC address as that assigned to the server of system 1 is detected on the bus 3 on the bus. .

Therefore, the communication is disabled, the initialization is stopped, the sleep process for waiting for a certain period is started, and after the sleep process for the certain period, the initialization is tried again and the bus 3 is reset.
As long as the same MAC address as its own MAC address is detected above, the first system server does not start communication.

In this way, only one server can communicate.

FIG. 3 is a diagram for explaining means for enabling communication with another system when an error occurs in a server during communication in the configuration of FIG.

Since the configuration of FIG. 3 is exactly the same as the configuration of FIG. 2, the description of the configuration is omitted to avoid duplication.

Now, it is assumed that a communication error has occurred in the system 0 server during communication with the client 2 for some reason.
Upon detecting a communication abnormality, the system 0 server deactivates its own LAN card and enters a sleep state for a certain period.

During this time, if the server of the first system attempts to start communication and tries to initialize, the server of the zero system stops communication and detects the same MAC address as its own MAC address on the bus 3. Never. Therefore, the first system server can start communication after determining that communication is possible, setting its own MAC address and IP address, and confirming communication with the client 2.

Even if the server of system 0 attempts to initialize after a sleep state for a certain period of time, the server of system 1 has already started communication, so the server 3 has the same MAC address on the bus 3 as its own. Of the bus 3
The sleep process and the initialization process for a certain period are repeated until the same MAC address as the own MAC address is not detected above.

In this way, when the server in communication becomes abnormal, the server of the other system becomes communicable, and thereafter, only the server of one system can continue the communication.

FIG. 4 is a diagram for explaining a control process in the present invention.

In FIG. 4, reference numeral 1 denotes a server.

Reference numeral 111 denotes a LAN control unit provided in the server 1, reference numeral 112 denotes a driver provided in the server 1, reference numeral 113 denotes a LAN card provided in the server 1, and reference numeral 113 denotes a LAN card provided in the server 1. The communication control unit 11 and the communication control unit 11a of FIG. 3 are configured by the driver 112 and the LAN card 113. Further, the driver 11
2 is provided with a MAC address management unit 112-1.

Although FIG. 4 shows the configuration of the server 1 of FIG. 3 for the time being, the configuration of the server 1a of FIG. 3 is the same.

Reference numeral 3 denotes a bus.

The driver 112 detects the MAC address flowing on the bus 3 via the LAN card,
The detected MAC address is stored in the MAC address management unit 11
2-1.

Here, the search of the MAC address will be briefly described.
An AC frame is being transmitted. Since the source MAC address is mounted at a fixed position in the MAC frame, the server extracts the source MAC address from all the MAC frames transmitted via the LAN card and is provided in the driver. Stored in the MAC address management unit.

When a server in such a state tries to start communication, it is sufficient to first access the MAC address management unit and determine whether or not the same address as its own MAC address is stored.

That is, the LAN control unit 111 searches the MAC address management unit 112-1 when performing initialization to start communication, and finds the same MAC address as its own MAC address (in the figure, MAC-A is abbreviated, but it means the MAC address. In the following description, the MAC address may be indicated in the same manner.), And a check result is received.

If it is determined from the received check result that the same MAC address as the own MAC address is not detected on the bus 3, the own IP address is set and the client is confirmed for communication.

When communication with the client is confirmed, the initialization is terminated and communication with the client is started.

On the other hand, when it is determined from the received check result that the same MAC address as that of itself is flowing on the bus 3, initialization is stopped and sleep processing for a certain period is started. In addition, when communication confirmation is not obtained, own LAN
The card is deactivated, and a sleep process is started for a certain period.

FIG. 5 is a flowchart for explaining the operation of the server of the present invention, which is shown in order to more clearly show the procedure described above with reference to FIG.

Hereinafter, the operation of the server of the present invention will be described along the reference numerals in FIG.

S1. The LAN control unit starts initialization of the LAN card.

S2. Search for the MAC address management unit; S3. Investigate the presence or absence of the same MAC address as its own MAC address.

S4. If it is determined in step S3 that the MAC address management unit has detected the same address as its own MAC address (Yes), the initialization process is stopped, and S5. After entering the sleep process for a certain period, the process returns to step S1 after a certain period.

S6. On the other hand, in step S3, MA
If the C address management unit determines that the same address as its own MAC address has not been detected (No), it activates its own LAN card.

S7. The LAN card has a TCP / IP address (in the figure, it is abbreviated as TCP / IP-A,
P / IP address. ) Is set.

S8. TCP / IP protocol (in the figure,
Although abbreviated as TCP / IP-P, it is a TCP / IP protocol. ) To confirm communication with the client.

S9. It is determined whether or not the communication with the client has been confirmed.

S10. If it is determined that communication confirmation with the client could not be performed (No), the own LA
The N card is deactivated, the processing is terminated, and the routine jumps to step S5.

S11. On the other hand, if it is determined in step S9 that the communication with the client has been confirmed (Y
In es), communication with the client is started.

S12. It is determined whether communication with the client has been completed.

When it is determined that the communication has not been completed (N
In o), the process jumps to step S9 and repeats steps S9, S11 and S1 until the communication ends.
Repeat 2.

S13. If it is determined in step S12 that the communication has been completed (Yes), the own LAN card is deactivated, and a series of processing ends.

With the above processing, only one of the redundant servers can communicate with the client.

FIG. 6 is a diagram for explaining a procedure for enabling communication of only one server in the configuration of FIG. 1. FIG. 6 shows the operation between a duplicated server and a duplicated server in a single client client server system. It is illustrated for the purpose of showing the relationship. Therefore, the operation itself of each server is omitted from FIG.

In FIG. 6, first, the 0-system server starts communication with the client, and then if the 1-system server tries to communicate, the 0-system server cannot communicate during the communication. This shows a situation in which the first system server communicates with the client after an error has occurred.

The server of system 0 performs the initialization process, and if the initialization is successful in the procedure described in the flowchart of FIG. 5, it allocates its own IP address and starts communication with the client via the bus. That is, the client communicates with the system 0 server.

Thereafter, the server of system 1 attempts to initialize to start communication. However, since the MAC address of the server of system 0 flows on the bus, the initialization fails and the sleep process starts.

If a communication error occurs between the server of the system 0 and the client for some reason and the communication between the server of the system 0 and the client cannot be confirmed, the communication of the system 0 is stopped. The server deactivates its LAN card and enters sleep processing.

If the server of the system 1 tries to initialize again while the server of the system 0 is in the sleep process, the server of the system 0 is in the sleep state at this time and the MAC address is stored on the bus. , The server of the first system succeeds in initialization, and can set its own IP address and start communication with the client.

Even if the server of the system 0 tries to initialize again after the server of the system 1 starts communication with the client,
At this time, since the first system server sends the MAC address on the bus, initialization fails and communication cannot be started.

In this way, when a communication error occurs between one of the redundant servers and the client, the other server can start communication with the client, and
At any given time, only one of the redundant servers can communicate with the client.

In the description with reference to FIGS. 5 and 6, it is assumed that the sleep processing period is constant, but the sleep processing period does not need to be constant.

For example, if the initialization is not successful even after performing the sleep processing a specified number of times, it means that another server has continued communication for a long time, and it is useless to repeat the sleep processing frequently. Therefore, it is also an effective technique to extend the period of the sleep process when the continuous number of sleep processes exceeds the specified number, or to extend the period each time the sleep process is performed.

In these, the number of sleep processes is counted, an offset value is read from a read-only memory (a so-called ROM) using the count result as an address, and the offset value is loaded into a counter for determining a sleep process period. And carry out at a predetermined count value. Since this technique is easy for those skilled in the art, illustration and further description are omitted.

Further, it is not essential to try the initialization again after the sleep process when the initialization is unsuccessful.
If the initialization is unsuccessful, it is possible to immediately shift to the stop state.

Since this can also be easily realized by those skilled in the art, further detailed description of the means will be omitted.

Normally, the hardware of the server or the client is often constituted by a complementary metal oxide semiconductor integrated circuit (a so-called CMOS-IC), so that the sleep processing is performed as described above. By extending the period to reduce the number of times of the initialization processing, there is an advantage that the power consumption of the server and the client can be reduced gradually.

FIG. 7 shows a client server system of a duplex server and a plurality of clients.

In FIG. 7, reference numerals 1 and 1a denote redundant servers. Here, it is assumed that the server 1 is sometimes called the 0-system server and the server 1a is sometimes called the 1-system server.

Reference numerals 2, 2a and 2b are independent clients.

Reference numeral 3 denotes a bus, and each client communicates with one of the redundant servers via the bus 3.

Reference numeral 4 denotes a path by which the server 1 accesses each client via the bus 3; 4a, a path by which the server 1a accesses each client via the bus 3; A path for the client 2 to access the bus 3, a path 5 a for the client 2 a to access the bus 3, and a path 5 b for the client 2
b is a path for accessing the bus 3.

As in the configuration shown in FIG. 7, the technology of the present invention can be applied to a case where a plurality of independent clients are connected to a duplicated server via a bus.

That is, the same MAC address and IP address are given to the server 1 and the server 1a in FIG.
Client 2, client 2a and client 2
By giving independent MAC addresses and IP addresses to b, the communication system already described in the subsystem composed of the server 1, the server 1a and the client 2 can be realized. Similarly, the communication system described above can be realized in the sub system including the server 1, the server 1a, and the client 2a, and the sub system including the server 1, the server 1a, and the client 2b.

That is, the communication method described with reference to FIGS. 2 to 6 is not limited to a client server system composed of a duplicated server and a single client, but is provided with the same MAC address and IP address. Duplicated server, independent MAC address and IP
The present invention can also be applied to a client-server system including a plurality of clients assigned addresses.

Further, the multiplexing of the servers does not need to be limited to the duplication. Assuming that the same MAC address and IP address are given, it is possible to communicate with the client in exactly the same procedure as in the case of duplexing, even if the multiplexing is more than tripled.

However, in practice, if the servers are duplicated, the reliability of the client-server system can be kept sufficiently high. Therefore, it is practically necessary to use servers multiplexed more than three times. Absent.

In FIGS. 2 to 6, a case has been described in which communication between a duplicated server and a single client is performed by a server, but the technique of the present invention is limited to server-driven communication. It is not something to be done. The detailed operation will be described with reference to a flowchart. However, by setting different sleep processing time settings in a redundant server, client-led communication is also possible.

FIG. 8 is a flowchart for explaining client-initiated communication.

Hereinafter, how the client-initiated communication is performed along the reference numerals in FIG. 8 will be described. Before that, the configuration and function of the server and the client, which are presupposed, will be added.

That is, the configuration of the client is based on the premise that the client includes a LAN control unit, a driver provided with a MAC address management unit, and a LAN card, similarly to the configuration of the server shown in FIG.

Also, even when the server itself stops communication, it is assumed that the server can capture packets flowing on the bus via the LAN card and extract the MAC address from the captured packets. I have.

In this method, a packet is simply taken in synchronously, and a MAC mounted at a predetermined position in the packet is taken.
It can be easily understood that it can be easily realized because only the address is extracted.

Now, S21. The LAN control unit of the client starts initialization of the LAN card.

S22. Searching the MAC address management unit; S23. Check the presence or absence of the MAC address of the server.

S24. In step S23, the MAC
If the address management unit determines that the MAC address of the server has been detected (Yes), the initialization process is stopped, and S25. After the sleep process, the process returns to step S1.

S26. On the other hand, in step S23,
If the MAC address management unit determines that the MAC address of the server has not been detected (No), the client's own LAN card is activated.

S27. Set a TCP / IP address on the LAN card.

S28. The client sends a packet conforming to the TCP / IP protocol to the server.

On the other hand, the 0-system server and the 1-system server operate as follows.

First, the server of system 0 is as follows.

S41. Communication has been stopped.

S42. It is determined whether the MAC address of the client is flowing on the bus.

If it is determined that the MAC address of the client is flowing on the bus (No), step S4
Return to 1.

S43. If it is determined in step S42 that the MAC address of the client is flowing on the bus (Yes), a sleep process is started. The period of the sleep process in the 0-system server is τ ₀ .

S44. When the period τ ₀ has elapsed, the sleep process ends, and the system 0 server ends the initialization and transmits a communication confirmation to the client.

Similarly, the first server is as follows.

S41a. Communication has been stopped.

S42a. It is determined whether the MAC address of the client is flowing on the bus.

If it is determined that the MAC address of the client is flowing on the bus (No), step S4
Return to 1a.

S43a. If it is determined in step S42a that the MAC address of the client is flowing on the bus (Yes), a sleep process is started. Interval of the sleep process in the 1-system server of the tau _1.

S44a. When the period τ ₁ has elapsed, the sleep processing ends, and the system 1 server ends the initialization and transmits a communication confirmation to the client.

Now, it is assumed that the sleep processing period τ ₀ set for the system ₀ server is shorter than the sleep processing period τ ₁ set for the system 1 server.

Accordingly, the timing at which the server of the 0 system issues a communication confirmation to the client in step S44 is earlier than the timing of the server of the 1 system issuing a communication confirmation to the client in step S44a. The communication confirmation of is received first.
(However, here, it is assumed that the transmission delay time of the communication confirmation is equal. If the transmission delay time between the two servers and the client is different, the difference is equal to the difference in the server having the smaller transmission delay time. A delay circuit may be provided to substantially reduce the transmission delay time difference to zero.) S29. The client determines whether the communication has been confirmed.

In this case, since it is assumed that the client has received the communication confirmation of the 0-system server, the client can naturally confirm the communication.

In consideration of the fact that communication cannot be generally confirmed, if communication cannot be confirmed, the own LAN card is deactivated in step S30 and step S2 is performed.
It is necessary to prepare a routine for entering the sleep processing of No. 5.

S31. If it is determined in step S29 that communication has been confirmed with the 0-system server, communication with the 0-system server is started.

S45. Since communication can be confirmed with the 0-system server substantially simultaneously with step S31, S46. The communication with the client is started, and a series of processing ends.

It should be noted that, in general, considering the case where the communication confirmation cannot be obtained by the server of system 0, its own LA in step S47 is considered.
It is necessary to prepare a routine for deactivating the N card and ending a series of processing.

On the other hand, the first-system server, which has a longer sleep processing period than the zero-system server, confirms communication with the client after communication between the client and the zero-system server is confirmed. At this time, at least communication between the system 0 server and the client has been confirmed, and in some cases, the communication between the system 0 server and the client has been started. Can not.

Therefore, in S45a. The first system server determines that the communication confirmation is unsuccessful (No), and S47a. Deactivate its own LAN card and end a series of processing.

In general, it is necessary to prepare a routine for performing communication with the client in step S46a in consideration of the case where communication confirmation can be performed by the first system server.

As described above, by making the sleep processing periods of the duplicated servers different, it is possible to start communication under the initiative of the client when both of the duplicated servers are stopped.

If τ ₁ is set smaller than τ ₀ , the server of the first system and the client can start communication, and the server of the zero system cannot start communication.

In the above description, the client-initiated communication start in the client server system composed of a duplicated server and a single client has been described. However, a duplicated MAC with the same MAC address and IP address is given. In a client / server system composed of a server and a plurality of independent clients given independent MAC addresses and IP addresses, a similar client is also used in a subsystem composed of a duplicated server and one independent client. Initiative communication is possible. That is, the technique described with reference to FIG. 8 is generally applicable to a client-server system including a duplicated server and a plurality of independent clients.

As described above, it is not technically necessary to limit the multiplexing of the servers to the duplication, but practically, the servers may be duplicated.

The fact that the client-initiated communication is possible in the client-server system to which the present invention is applied means that data transfer is mostly small and local at night, which is constituted by the client-server system. This is advantageous in an office system that does not have such a function, in which only some of the functions of the server are stopped and the functions are stopped for power saving.

FIG. 9 shows the client / server of the present invention.
This is an example in which the system is applied as a monitoring system.

In FIG. 9, 1 and 1a are servers, 2 is a client, 3 is a bus, 4 and 4a are paths for accommodating the server 1 and the server 1a on the bus 3, and 5 is a client for the bus 3 This is the path to be accommodated.

The server 1 includes a communication control unit 11, a data collection processing unit 12, and a data transfer unit 13. Similarly, the server 1a includes a communication control unit 11a, a data collection processing unit 12a, and a data transfer unit. 13a. still,
The configuration of the communication control unit is as shown in FIG.

When viewed as a component of the monitoring system, the server 1 and the server 1a are so-called monitoring devices that collect and process monitoring data to monitor the monitoring target system, and the client 2 is a monitoring target. It is a so-called sensor device that collects system status and alarms.

In the monitoring system, the monitoring device periodically polls the sensor device, and the sensor device that receives the polling transfers the collected and stored data to the monitoring device after transferring the data to the monitoring device first. .

At this time, the server 1 and the server 1a as the monitoring devices communicate with the client 2 as the sensor device according to the above-described procedure.

By the way, in the client server system of the present invention, the same MAC address and the same IP address are assigned to the two servers as the monitoring devices, and the same MAC address as its own MAC address is provided on the bus. When it is not flowing, one server as the monitoring device can communicate.

Since the client-server system having this communication method is applied to the monitoring system, only one of the servers as the monitoring devices can communicate with the client as the sensor device, and the other cannot.

Therefore, the real time data at a certain time is not transferred to only one monitoring device, and the other monitoring device cannot have the data at the same time.

To avoid this situation, the data transfer unit is provided in the server which is a monitoring device.

In one monitoring device that has collected data from the sensor device, the data collection processing unit transfers the currently received data to the data transfer unit. The data transfer unit transfers the transferred data to the data transfer unit of the paired monitoring device.

In the monitoring device that has received the data transfer from the paired monitoring device, the data transfer unit checks the data stored in the data collection processing unit to confirm that the data is new data. Later, the data transferred from the paired monitoring device is transferred to its own data collection processing unit.

In this way, the duplicated monitoring devices can always share the same data,
The monitoring work in both monitoring devices can be performed smoothly.

In the configuration shown in FIG. 9, a transmission path for transferring data between the duplicated monitoring devices is required, but this is not a disadvantage because the data can be transferred without passing through the bus. Rather, if data is to be transferred via the bus, it is necessary to give a different MAC address to the duplicated server, and the bus will be occupied during the data transfer. The configuration can be said to be an advantageous configuration.

In the above description of the monitoring system, it has been described that the data transferred by the data transfer unit is compared with the data stored in the data collection processing unit. However, although data collation is preferable in order to improve the reliability of stored data, data collation is not essential since data to be transmitted is always new data if data is reliably transmitted. In this sense, it is possible to omit the collation function from the data transfer unit.

Finally, although the above description has been given of an example in which the transmission path between the server and the client is a bus type, the form of the transmission path need not be limited to the bus type at all. That is, the present invention is not limited to the form of the transmission path as long as the transmission path can support communication by the TCP / IP protocol.

Therefore, what is described as a bus in the above can be generally replaced with a transmission line.

[0167]

According to the present invention, a client server system in which a plurality of servers communicate with each other by connecting a single client which can recognize only a single IP address to the plurality of servers by a TCP / IP protocol, and A monitoring system to which a client-server system is applied can be realized.

As a result, the scale of the client server system can be reduced, and similarly, the scale of the monitoring system can be reduced.

Furthermore, if the technique of the present invention is applied, a plurality of servers and a single client having only a single IP address for the plurality of servers are connected by the TCP / IP protocol regardless of the server or the client. Communication is possible.

That is, according to the present invention, it is possible to realize a client-server system having a small system scale, high system reliability, and flexibility, and a monitoring system to which the client-server system is applied. Can be.

[Brief description of the drawings]

FIG. 1 shows a dual server, single client client server system to which the present invention is directed.

FIG. 2 is a view for explaining means for enabling only one server to communicate in the configuration of FIG. 1;

FIG. 3 is a diagram illustrating a unit that enables another system to communicate when a server in communication is abnormal in the configuration of FIG. 1;

FIG. 4 is a diagram illustrating a control process according to the present invention.

FIG. 5 is a flowchart illustrating the operation of the server of the present invention.

FIG. 6 is an exemplary view for explaining a procedure for enabling only one server to communicate in the configuration of FIG. 1;

FIG. 7 shows a redundant server, a client / server system of a plurality of clients.

FIG. 8 is a flowchart in the case where communication is started from a client by changing the sleep processing time of the redundant server.

FIG. 9 is a monitoring system to which the client server system of the present invention is applied.

FIG. 10 is a basic configuration of a client server system in which a server is duplicated.

FIG. 11 shows communication between a client and a server in the configuration of FIG. 10;

FIG. 12 is a diagram (part 1) for explaining a problem of the conventional technique in the configuration of FIG. 1;

FIG. 13 is a diagram (part 2) for explaining a problem of the conventional technique in the configuration of FIG. 1;

[Explanation of symbols]

 1 server (0 system server) 1a server (1 system server) 2 client 2a client 2b client 3 bus 4 path 4a path 5 path 5a path 5b path 11 communication control unit 11a communication control unit 12 data collection processing unit 12a data collection Processing unit 13 Data transfer unit 13a Data transfer unit 111 LAN control unit 112 Driver 113 LAN card 112-1 MAC address management unit

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Yoshiko Tatsumi 14-15 Higashi-Shirashima-cho, Naka-ku, Hiroshima City, Hiroshima Prefecture Inside Fujitsu China Communication Systems Co., Ltd. (72) Inventor Shinji Yamane Higashi-Shirashima-cho, Naka-ku, Hiroshima City, Hiroshima Prefecture No. 14-15 Fujitsu China Communication Systems Co., Ltd. F term (reference) 5B034 CC05 5B085 AA01 AA08 AC16 BG07 5B089 GA11 GA21 HB02 KA12 KB06 KC41 KE03 KF03 MC08 ME02 ME10 ME14 5K035 AA04 AA07 BB02 CC09 EE03 LL07

Claims (5)

[Claims] 1. A client-server system comprising a plurality of servers, a plurality of clients, and a transmission path accommodating the plurality of servers and the plurality of clients, wherein the plurality of servers have the same A MAC address and an IP address are assigned, and the plurality of clients have independent MAC addresses and IP addresses.
At the time of initialization when one of the plurality of servers starts communication, the same MAC address as its own MAC address is transmitted on the transmission path.
It is determined whether or not an address is flowing, and when it is determined that the same MAC address as its own MAC address is not flowing on the transmission path, initialization is completed and the communication device itself starts communication with the client. A client-server system characterized in that when it is determined that the same MAC address as its own MAC address is flowing on a transmission path, initialization is stopped. 2. The client server according to claim 1,
A client-server system, wherein when a MAC address identical to its own MAC address is flowing on the transmission path, initialization is stopped and sleep processing is started. 3. The client server according to claim 2, wherein
A client-server system, wherein a period of the sleep processing is set to a fixed time. 4. The client server according to claim 1,
The system, wherein the plurality of servers set different sleep processing periods, and when one of the plurality of clients transmits a packet to the plurality of servers that have stopped communication, the plurality of servers sleep. A client server that returns a communication confirmation to the client after shifting to processing.
system. 5. A monitoring system to which the client server system according to claim 1 or 4 is applied, wherein the plurality of servers are monitoring devices, and the plurality of clients are sensor devices. A data transfer unit for directly transferring data between the monitoring devices in the monitoring device, and a transmission line for directly transferring data between the monitoring devices between the data transfer units provided in the monitoring device. A monitoring system comprising: