JPH0950417A - Client ip address dynamic change system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reduce the damage of user by setting a host group in a file on the client side and contacting a user on the client side when all hosts can not perform communication. SOLUTION: In this system constituted of servers 7-9 for processing plural events, a file server 10 and the various kinds of communication equipments 13, based on a host group file 5 for registering the address of a processing server for processing the plural events set on the side of the client 1, an IP address dynamic change program 2 on the side of the client 1 transmits a packet to the server at the time of requesting connection, analyzes information by operation condition data from a communication program on the server side and the response time, decides a connection server opposite party corresponding to a selection standard 6 and automatically transmits a connection request to the server. Also, a display device 4 on the side of the client 1 for displaying a message when the server inside the host group is in an access impossible state is provided.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to load balancing by checking the operating status of the server operating environment from the client side in the processing distribution of a plurality of servers for processing the same kind of business in a large-scale CSS environment and when a network device failure occurs. The present invention relates to an optimal processing server selection determination method in flexible fault avoidance.

[0002]

2. Description of the Related Art Conventionally, as a communication path determination method, there is an optimum message transfer path determination method in the Internet environment as described in Japanese Patent Laid-Open No. 6-120944. In this route determination method, the response time with the gateway machine is recorded, and the machine having the shortest response time determines the other party as the optimum machine.

[0003]

However, in a CSS environment in a computer system that requires high-speed processing capability, a server system environment in which processing by a plurality of servers for processing the same work is distributed is required as the computer scale increases. Will be. It was necessary to flexibly implement load balancing and failure countermeasures in this environment.

The present invention has been made to solve the above problems.

It is an object of the present invention to distribute the load of a plurality of servers that handle the same type of business in a large-scale computer system in a CSS environment and select the optimum server as the server connection destination when a failure occurs when a connection request is made from the client side. , To provide a reduction in disaster recovery time for client-side users.

[0006]

In order to achieve the above object, the communication software on the client side is in a connectionless format except when in use, and the communication software on the client side makes a connection request to the server side at the time of use. A method to do is necessary. At this time, a health check is performed by sending and receiving operation check packets of communication devices up to the server and all servers in the host group, and the server in the optimal state for the user to work is automatically selected and a connection request is made. .

If the connection request is not responded to all the servers in the host group, the timer monitoring value on the client side disables the connection and notifies the user on the client side. As a result, it is possible to isolate the failure location in the server and the network, reduce the analysis time, and secure the hardware failure recovery time.

[0008]

[Operation] A host group is set in the file on the client side and the user on the client side is notified when all hosts cannot communicate. Further, as the system and the network expand, it is possible to secure the failure response time, limit the failure location, and reduce the damage to the user.

[0009]

An embodiment of the present invention will be described in detail below with reference to the drawings.

FIG. 1 illustrates the overall structure of this embodiment. In a system including servers (7 to 9) that process a plurality of events, a file server (10), and various communication devices (13), the addresses of the processing servers that process a plurality of events set on the client (1) side Based on the registered host group file (5), the IP address dynamic change program (2) on the client side sends a packet to the server at the time of connection request, the response time and the operating status from the communication program on the server side. A function that analyzes the information based on the data, determines the destination of the connected server according to the selection criteria (6), and automatically sends a connection request to the corresponding server, and displays a message when the server in the host group is inaccessible It is composed of a client side display device (4).

FIG. 2 shows an example of a table configuration for setting the host name (22) of the server that processes a plurality of events on the client side. The servers (7 to 9) that process a plurality of events are grouped together. The name (21) is grouped and defined.

In FIG. 3, a timer value (32) for setting the failure monitoring time and the response response time from the server,
It is composed of data such as CPU performance, which is a selection criterion (31) at the time of connection connection, and load information in the network, and is used as a counterpart server selection criterion.

In FIG. 4, the partner server address (41) information, the response time (42), and the server operating status information (4
3) shows a table configuration for storing the data, and after the packet is received, the data is stored in the in-program table (3).

In FIG. 5, the client IP according to the present invention is shown.
The operation of the dynamic change program (2) will be described. When the system starts, the information storage table (12) is created (51) in the memory, and when the user connects from the client (1) side to the server by designating the host group (21), the corresponding information table (12) is temporarily displayed. Clear (52).

After that, timer monitoring is started to measure the response time (53), and a status confirmation packet is sent to the set servers all at once (54). The data is stored (57) in the operation information table (55) until an over occurs.

If a time over occurs at this time (58), it is judged as a failure and the maximum value of the data is stored (5).
9).

The client IP dynamic change program is
A connection destination server is selected according to the selection criteria set based on this information (62), and a connection connection request is automatically executed (63).

Further, when there is no response to the server in the host group definition (when the value is the maximum value), it is determined as a failure on the client side or a failure in the network environment,
The error information is displayed (61) on the display device (4) on the client side.

[0019]

According to the present invention, L constituting the system is
It is possible to automatically reduce the influence on the user of a failure that occurs in the AN environment, distribute the load on a plurality of servers, and avoid troubles at the time of failure.

[Brief description of drawings]

FIG. 1 is a system configuration diagram of an embodiment.

FIG. 2 is a host grouping definition diagram according to an embodiment.

FIG. 3 is a setting diagram of host selection criteria and timer monitoring time according to an embodiment.

FIG. 4 is a configuration diagram of an operation status storage table according to an embodiment.

FIG. 5 is a flowchart of the operation contents of an IP dynamic change program according to an embodiment.

[Explanation of symbols]

5 ... Files for grouping and managing servers, 6 ...
Connection standard setting file, 7-9 host for event processing, 10 host for file management.

Claims (1)

[Claims] 1. In a large-scale CSS (Client Server System) environment connected to a LAN environment such as Ethernet,
The number of servers is increased in order to increase the processing capacity of the same type of work, and the same processing is performed by a plurality of servers for decentralization. At this time, if the server that requests the connection from the client through various network devices, the LAN environment failure, or the server is overloaded, the server that processes the same kind of work on the client side is registered as one host group. A method for dynamically changing a client IP address, which is characterized in that an optimum server is selected from among the servers and IP addresses set in advance and the address is dynamically set.