KR100233245B1 - Redundancy Control Method in High Speed Wireless Calling System - Google Patents

Abstract

1. TECHNICAL FIELD OF THE INVENTION

Redundancy Control Method in High Speed Wireless Calling System.

2. The technical problem to be solved by the invention

In the redundant high speed wireless calling system, the system can be operated without interruption of service through effective redundancy control.

3. Summary of Solution to Invention

Read each system status information, communication port status information, system alarm status information and switch mode information from common memory to check the switch mode of itself and the other party, communication port status and alarm information of self and other party and operate according to the status Set the standby mode.If both yourself and the other party are in operation mode, refer to the time information to set the system that has been in operation mode for a longer time, and if both yourself and the other party are in standby mode, set the high priority system to the operation mode. Set it.

4. Important uses of the invention

Used for high speed wireless calling system.

Description

Redundancy Control Method in High Speed Wireless Calling System

The present invention relates to a redundancy control method in a high speed wireless call system, and more particularly, to a redundancy control method for determining an operation / standby mode using respective state information in a high speed wireless call system having a redundant structure.

Since the first wireless calling service in Korea, many subscribers have been accommodated through the development of various additional services and the development of technology to improve the supply and reception rate of various terminals.

The conventional low speed wireless calling system performs call processing using a TDX-PS. In such a wireless calling system, the call success rate exceeds 99% according to statistics. Among them, call loss rate by subscriber matching device is about 0.03%. These statistics show that the switching network (TDX-PS) of the conventional radio calling system is quite stable in terms of call completion rate.

The conventional wireless calling system uses up to four Voice Mailing Systems (VMSs) to provide voice mailbox services. The voicemail system of the conventional wireless calling system is a system made for a specific purpose, and has a very low scalability or compatibility of the system, and it is difficult to add a new service. In addition, the voice mailbox system is unstable, so that the frequency of the failure of the subscriber's voice files is high, and there are inconveniences in terms of network management, such as the need to periodically organize the voice files by the operator (administrator).

Therefore, in order to improve the disadvantages of the conventional wireless calling system as described above, a new high speed wireless calling system (APS: Advanced Paging System) has been studied.

The high-speed wireless call system is configured in the form of a distributed system and provides a graphical operator interface. In addition, there are many advantages, such as scalability of the system and ease of service development, by using a general-purpose computer with an open structure.

This high-speed wireless calling system has a redundant structure to ensure the stability of the system due to the distributed structure.

Accordingly, an object of the present invention is to provide a redundancy control method in a high-speed wireless call system, which enables the system to operate without interruption of service through effective redundancy control in the redundant high-speed wireless call system.

1 is a diagram showing a redundant structure of a fast paging system to which the present invention is applied.

Figure 2 is a diagram showing the redundant state of the server to which the present invention is applied.

3 is a diagram showing the software structure of a server for redundancy control to which the present invention is applied.

4 is a flowchart of a redundancy control method according to the present invention;

5 is a control flowchart according to a process restart condition of a process management unit;

In order to achieve the above object, the present invention provides a redundancy control method applied to a high-speed wireless calling system, which reads each system state information, communication port state information, system alarm state information, and switch mode information from a common memory. A first step of checking a switch mode of the other party; If the switch mode is manual, the current status is maintained. If the counterpart mode is manual, the next status is set to standby if the other system is in operation mode, and the next status is operated if the other system is in standby mode. Setting a second mode; A third step of checking a communication port state and alarm information of the self and the counterpart and setting an operation / standby mode according to the state; A fourth step of setting the system, which has been operated in the operation mode for a longer time, to the operation mode by referring to the time information when both the self and the counterpart are in the operation mode; A fifth step of setting a high priority system to an operation mode when both the self and the counterpart are in the standby mode; And a sixth step of setting the operation / standby mode, determining a set mode after a predetermined time has elapsed, and returning to an initial state.

Clients receive the caller's call request from the public telephone network and packetize it and send the packetized call message to the server using the TCP / IP protocol. In this case, in order to guarantee the quality of the call service, the path between the client and the server must be duplicated, and the server must be duplicated to increase availability. The present invention establishes a client / server network of a high-speed wireless paging system so that even if a failure occurs on one path, it can be transferred to another path and processed to ensure the quality of service through its redundant control.

Hereinafter, with reference to the accompanying drawings will be described an embodiment of the present invention;

1 is a diagram showing a redundant structure of a fast paging system to which the present invention is applied.

Although only two clients are shown in the figure, this may be further increased depending on the number of ports of a local area network (LAN) switch. In the drawings, 1 to 4 are links for connecting the LAN card and the LAN switch of the client, and if the reference numerals 1 and 2 are normal, any one of the higher priority is used for communication. When one or two of the lines fail, communication is performed through the non-failed line. However, if both 1 and 2 fail, there is no communication path between the client 1 and the server. Lines 3 and 4 connecting the client 2 and the server also have the same correlation as those of the lines 1 and 2 of the client 1.

If LAN card 5 fails, the client and server cannot communicate with each other via lines 2 and 4, in which case it is connected to LAN card 6 via lines 1 and 3, and LAN card 6 is immediately connected to ports 15 and 17 of the server. Connected with

If LAN card 6 fails, the client and server cannot communicate with each other via lines 1 and 3, which in turn connects to LAN card 5 via lines 2 and 4, which connects to the router via line 9, Is connected to Server 1 and Server 2 through 10 and 11.

Failures on lines 10 and 11 that connect the server to the router are detected by the server.

It is connected to the LAN switch via lines 12 and 13, and then directly to Client 1 and Client 2 using lines 1 and 3 or lines 2 and 4.

In the event of a failure in lines 12 and 13 connecting the server and the LAN switch, the server detects this and connects it to the router via lines 10 and 11, and then to the LAN switch via the connection of line 9 and lines 1 and 3 Or lines 2 and 4 to connect to Client 1 and Client 2, respectively.

As can be seen from the above, even if a failure occurs in one of the paths between the client and the server, the communication is performed through the other path.

2 is a diagram illustrating a redundant state of a server to which the present invention is applied.

In the figure, 25 and 26 represent respective servers, which are connected in a redundant structure. That is, each server monitors each other, and one server is determined as an active mode and the other server is determined as a standby mode so that only a server operating in the operational mode can provide a service. In case of failure of the server operating in the operation mode, the server in standby mode is changed to the operation mode to provide the service.

In the figure, subnets 20 and 21 are divided by router 19, and subnet 20 corresponds to those connected by the LAN switch of FIG. Therefore, all clients are connected to subnet 1. At this time, the LAN switch should be duplexed with the control unit and the power supply, and the router 19 should also be duplicated with the power supply and the control unit.

The above-described configuration is just an example, and it is possible to connect clients by connecting a LAN switch to subnet 2 as well. If possible, it is desirable for the client to have two LAN cards, each connected to a LAN switch on Subnet 1 and Subnet 2, to achieve redundant paths. It is also desirable to connect more subnets to duplicate paths.

Subnet 1 and subnet 2 are both networks that provide services and are also used for mutual monitoring between two servers 25 and 26. When two LAN cards of the operation mode server fail, communication with the outside is lost. At this time, the service in the standby mode must recognize the service and provide the service. However, if one LAN card is normal, communication with the outside is possible, so that service can be continuously provided. In particular, not only communication between 21, 23, 22, and 24 belonging to the same subnet is possible, but also communication between 21, 23, 22, and 24 belonging to different subnets allows two servers to operate at least one LAN card normally. If you do, you can monitor each other. If one server or two servers fail in both LAN cards, the communication between each other is lost and the state cannot be monitored, and it may take a long time to determine the status when the communication resumes. In order to ensure communication between the two 18 additional serial lines to provide information to each other in the event of a failure.

3 shows a software structure of a server for redundancy control to which the present invention is applied.

The alarm monitoring unit 31 monitors the state of the server excluding the LAN card and records in the common memory 32 when a serious failure affecting the service occurs. LAN watchdogs (29, 33) located in each of the two servers periodically send and receive information on the operating / standby status, alarm status, and LAN card status with the other party through normal LAN cards, and their next operational / standby status. Determine. The serial port external docks 30 and 34 located in each of the two servers allow information to be exchanged between servers when the communication between the servers is not possible due to an error in the LAN card. Operation / standby status determined based on the information of the partner server transmitted by the LAN watchdog (29, 33) or serial port watchdog (30, 34), the status of the LAN card monitored by the LAN watchdog, and various information. Is stored in the common memory 32. According to the operation / standby state stored in the common memory 32, the process manager 28 determines the process to be started or terminated and manages the processes.

All information related to the system is stored in common memory. Information stored in this common memory should consider the following time constraints. The two servers periodically send and receive messages over a LAN or serial line. This period must take into account the transmission speed of the LAN or serial line. In particular, when the service and the status monitoring is performed at the same time in the LAN, care should be taken because the performance of the service may be degraded due to the status monitoring message. The size of the mutual watch message is arbitrary m bytes, and the frequency at which the message is sent is arbitrary. T _periodic In seconds, the traffic required for health monitoring messages is m / T _periodic Bytes / sec. Send and receive messages T _periodic Sets the traffic required for the status monitoring message to be negligible (about 10000) compared to the traffic required for the service.

If the status change information of the counterpart server changes from time to time, the status change of itself is also frequent, so it is important for the stability of the whole system not to change the status until it is stabilized to some extent after one status change. Time to ignore this state change of other party T _ignore Cycles messages T _periodic It must be much larger than if T _ignore end T _periodic If it is smaller or similar, that is, it is smaller or similar to the period of sending and receiving messages, it is not preferable because the state change information of the other party is almost all reflected in the state decision.

When a LAN card failure occurs at the other server, messages cannot be received through the same subnet. T _forcedown If there is no response during this time, it is regarded as a fault and changes state. When the opponent is in the operation mode, the average transfer time for the player to enter the operation mode due to the obstacle of the opponent is T _forcedown - T _periodic / 2 + T _{recog-periodic} / 2.

T _periodic and T _{recog-periodic} If the time difference is large, the message may be accumulated on the sending or receiving side, so it should be set similarly. therefore, T _forcedown Is almost equal to the switching time of the system.

In consideration of the above conditions, the next operation / standby state of the server should be determined based on the information stored in the common memory, which will be described with reference to the flowchart of FIG. 4.

4 shows a flowchart of a redundancy control method according to the present invention.

Information related to the status change includes the partner server, its switch mode, LAN card status, operation / standby status, and system-wide alarm information. In switch mode, there are automatic / manual, manual means that the specified state is maintained and automatic means that the next state is determined by the law of change.

In addition to determining the status of redundancy, it is necessary to work to ensure that the status of the operation mode server and the standby mode server does not interfere with the continuity of services. One is to identify the database, and the other is to match the state of the queue. Since the contents of the database should be managed consistently, the process to ensure that the database modification request for the production server is transmitted to the standby server so that the changes made on the production server are the same on the standby server. The other is to manage the state of the queue, which periodically writes the contents of the changed queue to the disk on the standby server. If a switchover occurs, the server's process, which was the standby server, reads the information from the disk and writes it to its queue to recover.

A determination process of the duplex operation mode will be described with reference to FIG.

The state of the system, the LAN card state, the alarm state, and the switch mode are read from the common memory (101) to confirm whether the switch mode thereof is automatic (102).

If the switch mode is not automatic, the current state is maintained (103). If the switch mode is automatic, it is checked whether the mode of the counterpart switch is automatic (104).

If the counterpart switch mode is not automatic, it is checked whether the counterpart system is in operation mode (105). If the counterpart system is not in the operation mode, the next state is set to the operation mode (106), and if the counterpart system is in the operation mode, the next state is set to the standby mode (107).

If the other party's switch mode is automatic (104), it checks whether its LAN card is disabled (108). If its LAN card has failed, it sets its next state to standby mode (109). If one's LAN card is not a failure and the other party's LAN card is a failure (110), the next state is set to one's own operation mode (111). If the other party's LAN card is not a failure, if the alarm state of the own system is normal and the other party's alarm state is abnormal (112), the next state of itself is set to the operation mode (113). On the contrary, if the alarm state of the own system is abnormal and the counterpart alarm state is normal (114), the next state of itself is set to the standby mode (115). If the user is also in the operation mode and the other party is also in the operation mode (116), the next state is set to R1, where R1 is determined by referring to the time information to determine the server which has been in the operation mode for a longer time. (117).

On the contrary, if the user is in the standby mode and the other party is also in the standby mode (118), the next state is set to R2, where R2 is to determine the operation mode by giving priority to one server (119).

If the user is in the operation mode and the other party is in the standby mode (120), the next state of the user is set to the operation mode (121). Set to the standby mode (123). When the decision wait period ends (124), the process of reading status information from the common memory is repeated.

5 is a control flowchart according to a process restart condition of the process management unit.

If the process should check the restart condition of the process (NEVER), it checks the current process status and forcibly terminates (DO_KILL) if it is running (201, 202, 203). No take (DO_NOTHING) 204.

If the restart condition of the process must always be alive (ALLWAYS), the process state is taken over and no action is taken if it is in operation (205, 206), and if the process is stopped, the process is started (DO_ FORK; 207).

If the restart condition of the process operates only in the standby mode (Standby only), the system state is checked and if the process restarts, the process state is checked again (208, 209). If the process state is running, it is forcibly terminated (210). If it is stopped, no action is taken (211).

If the system state is in standby mode (208), the process state is checked and no action is taken if in operation (212, 213), and if it is in a stopped state, the process is restarted (214).

If the restart condition of the process is operated only in the operation mode (Active only), the system state is checked and if the operation mode is in the operation mode, the process state is checked again (215, 216). If the process state is running, no action is taken (217), and if the process state is stopped, restarted (218).

If the system state is in the standby mode, the process state is checked, and if the process state is in operation, it is forcedly terminated (215, 219, 220). If the state is stopped, no action is taken (221).

The present invention described above is not limited to the above-described embodiment and the accompanying drawings, and various substitutions, modifications, and alterations are possible within the scope without departing from the technical spirit of the present invention. It will be clear to those who have knowledge of

The present invention made as described above can effectively operate the system without interruption of service through the redundant control in the redundant high-speed wireless call system, it is possible to improve the availability and reliability of the system.

Claims (1)

In the redundant control method applied to a high speed wireless calling system, A first step of reading the respective system status information, communication port status information, alarm status information of the system, and switch mode information from the common memory to confirm the switch mode of the self and the counterpart; If the switch mode is manual, the current status is maintained. If the counterpart mode is manual, the next status is set to standby if the other system is in operation mode, and the next status is operated if the other system is in standby mode. Setting a second mode; A third step of checking a communication port state and alarm information of the self and the counterpart and setting an operation / standby mode according to the state; A fourth step of setting the system, which has been operated in the operation mode for a longer time, to the operation mode by referring to the time information when both the self and the counterpart are in the operation mode; A fifth step of setting a high priority system to an operation mode when both the self and the counterpart are in the standby mode; And And a sixth step of determining the set mode after a predetermined time elapses after setting the operation / standby mode, and returning to an initial state.

Images