JP2001268094A - Mode unit, information processing system provided with the mode units, and parameter exchange method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To realize parameter information exchange between its own node unit and another node unit on a network without increasing its own load and a network load. SOLUTION: A parameter storage area specific to each of computers 2-1-2-n is assigned to a shared memory 22-1 (i=1-n) on a computer 2-i connected to the network 1. A shared memory access section 24-1 writes various parameters generated by its own computer to a parameter storage area specific to its own computer in the shared memory 22-i and broadcasts the various parameters generated by its own computer to all the other computers via the network 1. When a parameter reception section 26-1 of the computer 2-i receives the various parameters broadcast from the other computers, on the other hand, the reception section 26-i updates contents of a parameter storage area specific to the other computers in the shared memory 22-i with the received parameter.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a node device suitable for sharing various parameter information generated by each of a plurality of node devices connected to a network among the node devices, and information provided with a plurality of such devices. The present invention relates to a processing system and a parameter exchange method.

[0002]

2. Description of the Related Art Conventionally, various information processing systems have been developed in which a plurality of information processing apparatuses (for example, computers) are interconnected by a network, and information is exchanged between the information processing apparatuses via the network.

[0003] Some information processing systems of this type exchange various parameters (parameter information) including statuses and the like generated by each information processing apparatus. For example, there is a complex computer system that exchanges parameters between an active computer and a standby computer that perform the same processing.

Conventionally, in order to exchange various parameters between information processing devices (hereinafter, referred to as node devices) connected to a network, the node devices on the network are connected to other node devices (or other node devices in the same group). In general, a method of individually transferring its own parameter is applied to each node device.

[0005]

As described above, in a conventional information processing system in which a plurality of node devices are interconnected by a network, in order to exchange various parameters between the node devices, the node devices must It has been general to apply a method of individually transferring its own parameters to other node devices (or other node devices in the same group).

However, in the conventional method, there is a problem that the load on each node device is large and the network has excessive traffic (that is, the network load increases). This problem became more remarkable as the number of node devices increased.

The present invention has been made in view of the above circumstances, and an object of the present invention is to exchange parameter information with another node device on a network without increasing its own load and the network load. It is an object of the present invention to provide a node device, an information processing system including a plurality of the devices, and a parameter exchange method.

[0008]

SUMMARY OF THE INVENTION The present invention relates to a node device applied to a system for exchanging various parameter information generated by each of a plurality of node devices connected to a network between the node devices. A shared memory in which a unique parameter storage area is reserved for each of the plurality of node devices including the device, and various parameter information generated by the own node device are written in the parameter memory area unique to the own node device in the shared memory. Shared memory access means, parameter transmitting means for broadcasting or multicast transmitting various parameter information generated by the own node device to the plurality of node devices other than the own node device via a network, and broadcasting or multicast transmission from another node device Various parameter information Receiving and, characterized by comprising a parameter receiving unit configured to update the contents of the specific parameter storage area by the received parameter information to the other node devices in the shared memory.

In the node device having such a configuration, its own parameter information can be broadcast or multicast transmitted to all the other node devices requiring the parameter information. Load and network load can be reduced. In addition, since the parameter information of each node device including the parameter information of the own node device is written in the parameter storage area unique to the node device in the shared memory owned by itself, various types of the plurality of node devices are stored on the shared memory. If parameter reference means for referring to the parameter information is provided, various parameter information of a plurality of node devices including the own node device can be obtained only by accessing the shared memory of the own node device. This means that the parameter information exchange with other node devices on the network can be realized without increasing the load on the network and the load on the network.

Here, the parameter transmitting means may have the following function, that is, a function of adding identification information of the own node device to the parameter information to be transmitted. In this case, when the parameter information is broadcast or multicast transmitted, the parameter receiving means on the node device side that has received the parameter information uses the node device identification information attached to the parameter information to read the parameter information on the shared memory. The parameter storage area to be updated can be easily specified.

The parameter information to be transmitted may be obtained from a parameter storage area unique to the own node device on the shared memory. Further, if a configuration is used in which each parameter information is transmitted in a unit of a predetermined parameter information group at a period unique to the parameter information group, the load on the node device and the network load can be further reduced. it can. As the unit of the parameter information group, each process of generating the corresponding parameter information group, or each related parameter type can be applied. Here, parameter information requiring urgency or parameter information whose status changes frequently may be transmitted in a shorter cycle, and parameter information that is not so may be transmitted in a longer cycle.

Further, according to the present invention, a plurality of node devices having the above configuration are connected to a network, and various parameter information generated by each of the plurality of node devices can be exchanged between the node devices via a shared memory. It is also characterized by a possible information processing system. Here, if the network is a local network and a system other than the plurality of node devices is not connected to the network, parameter transmission means of each node device may perform parameter transmission by broadcast. In a local network, a system in which node devices other than the plurality of node devices are connected, or a system in which an unspecified number of various node devices are connected to the network, such as the Internet, the plurality of node devices are specified. What is necessary is just to perform multicast transmission which is grouped into a group and simultaneously transmitted only to each node device in the group.

Further, the invention relating to the node device having the above configuration is also realized as an invention relating to a parameter exchange method for exchanging various parameter information generated in each of a plurality of node devices connected to the network between the node devices. I do.

[0014]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing a configuration of a computer system (information processing system) according to an embodiment of the present invention.

In FIG. 1, a network, for example, a local network 1 is connected to n node devices, for example, n computers (computers) 2-1 to 21-n. Computer 2-1 through network 1
For communication between 21-n, TCP / IP (Transmission Control
A communication protocol (TCP / IP protocol) called ol Protocol / Internet Protocol is applied.

Each computer 2-i (i = 1 to n) includes a CPU 21-i serving as a control center of the computer 2-i.
A shared memory 22-i for sharing parameter information generated by each of the computers 2-1 to 2-n;
And a communication interface (hereinafter, referred to as a communication IF) 23-i for communicating with another computer via the communication interface.

Each computer 2-i also retrieves parameter information of its own computer from the shared memory 22-i and transmits it by the communication IF 23-1 (broadcast transmission here), and a network from another computer to the network. And a parameter receiving unit 26-i for receiving parameter information transmitted (in this case, broadcast transmission) from the communication IF 23-1 via the communication interface 23-1 and storing the parameter information in the shared memory 22-i. A shared memory access unit 24-i for writing / reading information. Shared memory access unit 24-i
Are various processes operating on the CPU 21-i (here, for the sake of simplicity, the processes Ai and Bi
2) and the shared memory access unit 24-i.

FIG. 2 shows an example of area allocation in the shared memory 22-i. First, the area of the shared memory 22-i is divided into n and assigned to each of the computers 2-1 to 2-n. Identification information of each of the computers 2-1 to 2-n (computer I
D) is ID1 to IDn. Therefore, each computer 2
The area allocated to -1 to 2-n is defined as the ID1 area 220-.
1 to ID1 area 220-n. ID1 area 220-1
Are managed by being divided into areas in which parameter information unique to processes operating on the CPU 21-1 of the corresponding computer 2-1 is stored. Here, since there are two processes that operate on the CPU 21-1: A1 and B1,
The D1 area 220-1 is managed by being divided into a process A1 area 221A and a process B1 area 221B. FIG.
5 shows a state where m pieces of parameter information PAR1 to PARm are stored in the process A1 area 221A. It should be noted that the ID1 area 220-n and other ID areas unique to other computers are also included in the ID1 area 22-n.
Same as 0-1.

Next, the operation of the computer system having the configuration shown in FIG.
The description will be made with reference to the flowcharts of FIG. 4 and FIG. In a computer on the network 1, for example, the computer 2-1, processes A1 and B1 operating on the CPU 21-1 each generate a predetermined parameter group. These parameters are roughly classified into parameters (status parameters) indicating various statuses relating to processing performed by the corresponding process, and parameters (command parameters) indicating control commands. The command parameter itself indicates the content of the control (command).

The shared memory access unit 24-1 accesses the shared memory 22-1 and, based on a group of parameters generated in the processes A1 and B1, an ID unique to the processes A1 and B1.
The corresponding parameter group in the process A1 area 221A and the process B1 area 221B in the 1 area 220-1 is updated. The above operation is similarly performed on another computer such as the computer 2-n.

On the other hand, the parameter transmission unit 25-1 of the computer 2-1 is configured to execute each process A running on the CPU 21-1.
The shared memory 22-1 is accessed at a period (time interval) determined by each of the processes A1 and B2, and an I
The corresponding parameter group is read from the process A1 region 221A and the process B2 region 221B in the D1 region 220-1 (steps S1 and S2).

The parameter transmitting unit 25-1 transmits a broadcast frame (broadcast transmission frame) for broadcast transmitting the read parameter group to all the computers on the network 1 except the computer 2-1 each time the parameter group is read. (Step S
3), the data is transmitted over the network 1 by the communication IF 23-1 (step S4).

As shown in FIG. 3A, the broadcast transmission frame includes an IP header 31 and a UDP (Us
er Datagram Protocol) header 32 and a data section 33. This data section 33 includes the above-described step S3
In, a parameter group (parameter data group) 331 to be transmitted is added with a process identifier (process ID) 332 indicating a process corresponding to the parameter group and an identifier (computer ID) 333 of the own computer. Here, A1 (indicating process A1) is added as process ID 332, and ID1 (indicating computer 2-1) is added as computer ID 333.

If the size of each parameter is not constant, size information may be added to each parameter so that each parameter can be easily separated from the data section 33 on the receiving side. If the parameter transmission cycle is determined not for each process but for each parameter (parameter type), a parameter identifier (parameter number) is assigned to each parameter so that each parameter can be identified on the receiving side. In addition, a corresponding parameter identifier (parameter number) may be added to a parameter to be transmitted.

The IP header 31 has a destination IP address 310 as shown in FIG. In this embodiment, a broadcast address 311 is used as the destination IP address 310 as shown in FIG.
The UDP header 32 has a well-known source port number 321 and a destination port number 322, as shown in FIG. As the port numbers 321 and 322, port numbers (indicating the application programs) for exchanging data with an application program (application layer) that handles parameter information are used.

Now, the communication IF 23-1 of the computer 2-1
The transmission frame (broadcast transmission frame) of the format shown in FIG. 3 transmitted from the computer 1 to the network 1 is received by the communication IFs of all the computers other than the computer 2-1 on the network 1. Then, a computer other than the computer 2-1 such as the computer 2-n performs the following operation.

First, the communication IF 23-n of the computer 2-n
I of the frame (from computer 2-1) received at
The broadcast address 311 is used as the destination IP address 310 in the P header 31. For this reason,
The frame received by the communication IF 23-n is taken into the communication IF 23-n without being discarded. As the destination port number 322 of the UDP header 32 of the received frame, a port number indicating an application program that handles parameter information is used. Therefore, the data section 33 of the received frame is passed to the parameter receiving section 26-n in the computer 2-n.

The parameter receiving unit 26-n specifies the source computer and the process from the computer ID 333 and the process ID 332 in the received data (data of the data unit 33) (step S11). Here, since the computer ID = ID1 and the process ID = A1, it is specified that the source computer and the process are the computer 2-1 and the process A1.

Next, the parameter receiving unit 26-n accesses the shared memory 22-n, and stores the parameter group stored in the process area unique to the computer and the process specified in step S11 into the reception data (data (The data of the section 33) (step S1).
2). In this case, since the computer 2-1 and the process A1 are specified, the I-2 unique to the computer 2-1 is specified.
Process A unique to process A1 in D1 area 220-1
The parameter group in the 1 area 221A is updated to the latest state.

The above-described operation at the time of receiving the data broadcast-transmitted from the computer 2-1 is performed not only by the computer 2-n but also by all the computers other than the computer 2-1. As a result, the contents of the process A1 area 221A in the shared memories 22-1 to 22-n of all the computers 2-1 to 2-n on the network 1 are matched.

Similarly to the transmission of the parameter group corresponding to the process A1 by broadcasting, the computer 2-1 transmits the parameter group corresponding to the process B1 by broadcasting.
This is performed at a period unique to 1. Similar broadcast transmission is performed by other computers on the network 1 such as the computer 2-n.

As described above, each of the computers 2-1 to 2-n on the network 1 does not individually communicate with another computer, and the shared memory 22 of each of the computers 2-1 to 2-n does not communicate. Each ID1 area 220-j (j = 1 to n) (unique to the same computer 2-j) in -1 to 22-n
Can always be matched with the latest one. Therefore, the CPUs 21-1 to 21 of the computers 2-1 to 2-n are used.
Processes A1, B1 to An, and Bn operating on -n can refer to parameters of their own computer as well as parameters of other computers on shared memories 22-1 to 22-n of their own computer.

This means that each of the computers 2-1 to 2-n
This indicates that the above (processes A1, B1 to An, Bn) can exchange parameters with processes of other computers via their own shared memories 22-1 to 22-n. For this reason, in the case of a computer system that performs the same processing as an example, one of the computers 2-1 to 2-n is an active system and the other is a standby system.
2-n can monitor the status of another computer on the shared memories 22-1 to 22-n of its own computer,
Also, control can be easily performed.

Note that not all computers (node devices) connected to the network 1 need to exchange parameters. In particular, when the network 1 is not a local network such as the Internet, some of the computers connected to the network 1 include:
Generally, there are computers (node devices) that are not subject to parameter exchange.

In such a case, a group of computers for which parameters are to be exchanged (computer 2-1 in the example of FIG. 1)
2n), and using a multicast address 312 (instead of the broadcast address 311 in FIG. 3C) as the destination IP address 310 as shown in FIG. It is preferable to perform multicast transmission only to the (computers 2-1 to 2n). That is, the multicast address 31 including the identifier “1110” indicating the multicast communication (class D) and the group address
2 should be used. Note that the group address may be called a multicast address.

The case where the node device is a computer has been described above, but the present invention is not limited to this, and any node device that needs to exchange parameters via a network may be used.

[0037]

As described above in detail, according to the present invention, a unique parameter storage area is assigned to each of a plurality of node devices including the own node device in the shared memory on the own node device. While the generated various parameter information is written in the parameter storage area unique to the own node device in the shared memory, the various parameter information generated in the own node device is broadcast or multicast transmitted to other node devices via the network. hand,
Since the parameter information is written in the parameter storage area unique to the own node device in the shared memory of each of the other node devices, various parameter information of the plurality of node devices including the own node device is stored in the shared memory of the own node device. Can be referred to. As a result, parameter information exchange with another node device on the network can be realized without increasing its own load and the network load.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of a computer system according to an embodiment of the present invention.

FIG. 2 is a diagram showing an example of area allocation of a shared memory 22-i (i = 1 to n) in FIG. 1;

FIG. 3 is a diagram showing a format example of a broadcast transmission frame.

FIG. 4 is a flowchart for explaining parameter transmission processing by a parameter transmission unit 25-1 in the embodiment.

FIG. 5 is an exemplary flowchart for explaining processing at the time of parameter reception by the parameter receiving unit 26-n in the embodiment.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 ... Network 2-1 to 2-n ... Computer (node apparatus) 21-1 to 21-n ... CPU 22-1 to 22-n, 22-i ... Shared memory 24-1 to 24-n ... Shared memory access Units 25-1 to 25-n: Parameter transmitting unit 26-1 to 26-n: Parameter receiving unit 31: IP header 32: UDP header 33: Data unit 220-1 to 220-n: For ID (ID1 to IDn) Area 221A ... Process A1 area 221B ... Process B1 area 310 ... Destination IP address 311 ... Broadcast address 312 ... Multicast address 331 ... Parameter group 332 ... Process ID 333 ... Computer ID A1, B1-An, Bn ... Process (refer to parameters) means)

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat ゛ (Reference) // G06F 13/00 351 H04L 13/00 305Z 9A001 F term (Reference) 5B045 BB04 BB32 DD03 5B089 GA21 GB02 HA06 JB22 KA07 KB10 5B098 GC16 5K030 GA11 JT02 LD02 5K034 AA05 BB06 DD03 HH01 HH02 HH64 KK27 9A001 BB04 BB06 CC02

Claims (6)

[Claims] 1. A node device applied to a system for exchanging various parameter information generated by each of a plurality of node devices connected to a network between the node devices, wherein the plurality of node devices including its own node device A shared memory in which a unique parameter storage area is allocated to each of the shared memory devices; a shared memory access means for writing various parameter information generated by the own node device into a unique parameter storage region of the own node device in the shared memory; A parameter transmitting unit that broadcasts or multicasts various parameter information generated by the node device to the plurality of node devices except the own node device via the network; and transmits various parameter information broadcast or multicast transmitted from another node device. Receive Node device characterized by comprising a parameter receiving unit configured to update the contents of the specific parameter storage area by the received parameter information to the other node devices in the shared memory. 2. The apparatus according to claim 1, wherein the parameter transmitting unit adds the identification information of the own node device to the parameter information and transmits the parameter information. The parameter receiving unit adds the parameter information to the parameter information when receiving the parameter information from another node device. 2. The node device according to claim 1, wherein the parameter storage area to be updated on the shared memory is specified by the node device identification information. 3. The parameter transmitting means transmits various parameter information generated by the own node device for each predetermined parameter information group at a period peculiar to the predetermined parameter information group. 2. The node device according to 1. 4. The node device according to claim 1, further comprising parameter reference means for referring to various parameter information of said plurality of node devices on said shared memory. 5. An information processing system comprising: a network; and the plurality of node devices according to claim 1 connected to the network. 6. A parameter exchange method for exchanging various parameter information generated in each of a plurality of node devices connected to a network between the node devices, wherein the shared memory on the own node device includes the own node device. Each of the plurality of node devices secures a unique parameter storage area, and writes various parameter information generated by the own node device into the parameter storage area unique to the own node device in the shared memory. Broadcast or multicast transmission of the various parameter information generated in the plurality of node devices other than the own node device via the network, when receiving various parameter information broadcast or multicast transmitted from other node devices,
By updating the content of the parameter storage area unique to the other node device in the shared memory with the received parameter information, the latest of the plurality of node devices including the own node device on the shared memory of the own node device is updated. A parameter exchange method characterized in that various parameter information can be referred to.