KR100693663B1 - System and Method for detecting obstacle of node - Google Patents

Abstract

According to the present invention, when resource operation information for a node is received in a system configured with a heartbeat channel, the resource operation information is analyzed to determine whether there is a resource whose operation is stopped, and as a result of the determination, the resource whose operation is stopped is determined. If it exists, it is determined that the node has failed, and the failure information of the failed node is transmitted to another node and the service performed by the failed node is performed. It can also be used to quickly detect and overcome node failures to provide continuous service.

Node failure, heartbeat

Description

System and Method for detecting obstacle of node             

1 is a diagram showing the structure of a conventional information communication system.

2 is a flowchart illustrating a failure determination method of a conventional node.

3 is a diagram schematically showing the configuration of an information communication system for monitoring a node failure according to an embodiment of the present invention.

4 is a flowchart illustrating a method for changing configuration information on node resources according to an embodiment of the present invention.

5 is a flowchart illustrating a failure determination method of a node according to an exemplary embodiment of the present invention.

<Explanation of symbols for the main parts of the drawings>

300: node 310: management module

320: shared disk

The present invention relates to a node failure detection method and system that can quickly detect and overcome failures between nodes by inspecting whether a counterpart node is disabled using a control board of a node in a system environment configured with a heartbeat channel.

In general, an information communication system is composed of two or more multi-process systems, a shared disk, a heartbeat network and information communication software, and the information communication software analyzes the state of the system and transmits the information to another node. In the event of a failure, it is an important component that takes over the service that the node was performing to another node.

In addition, the information communication system is a system that recognizes the failure and takes over the service of the failing system and guarantees the continuity of the service when a hardware or software failure occurs in a system providing the service.

In recent years, as information technology advances, the performance and stability of hardware have been rapidly improved, but the software has provided a lot of obstacles due to the relatively increased complexity of the software.

Therefore, the ability to monitor and overcome software failures in information and communication systems is becoming increasingly important.

1 is a view showing the structure of a conventional information communication system.

1, an information communication system includes a plurality of nodes (100a, 100b, ... 100n, hereinafter 100), a heartbeat network for monitoring and exchanging information with each other. , A shared disk 120 for sharing data.

A method of detecting and processing a failure in the information communication system configured as described above will be described.

The information communication system recognizes a failure by transmitting and receiving information with each other through a heartbeat channel connected between the nodes 100.

The method for recognizing the failure of the node 100 through the commercial heartbeat channel is as follows.

First, there is a method using a network.

The above method is a method for checking whether a counterpart node is alive or a service is operating in the counterpart node by packets coming from the counterpart node through a network connected between the nodes 100.

When information transmitted from the first node to the counterpart node is transmitted in accordance with the format promised between the nodes, the other node receives the packet and checks whether a service running in the counterpart node is operating normally. As a result of the check, if the service being executed in the counterpart node does not operate normally, it recognizes that a failure has occurred and performs the service being performed in the counterpart node in its own node.

Next, there is a method using the shared disk 120.

The method of using the shared disk 120 is a method of checking the failure of the counterpart node using a device that the nodes 100 access and use in common. The shared disk 120 is used to recognize a failure of the counterpart node when communication over the network is not possible.

When the partner node is constantly in contact with the shared disk 120 when it is unable to collect the information of the partner node through the network, it is determined to be normal. An example of a device used for the shared disk 120 is a storage device such as a disk array or a tape library.

A failure determination method of a node performed by the above method will be described with reference to FIG. 2.

2 is a flowchart illustrating a failure determination method of a conventional node.

Referring to FIG. 2, when a failure occurs in a hang state of the first node (S200), the normal node does not receive a packet from the first node (S202). The healthy node then recognizes that the failed node does not use a shared disk.

When step 202 is performed, the normal node transmits a node resource operation information request command to the first node (S204). Here, the node resource may refer to a CPU, a memory, a disk, or the like.

Then, the normal node determines whether operation response information corresponding to the node resource operation information request command is received from the first node (S206).

If the operation response information is received from the first node as a result of the determination in step 206, the normal node determines that the first node is normal and does not provide the client with the service performed by the first node (S208). That is, when the operation response information is received from the first node, the normal node determines that the first node is normal and does not perform the service performed by the first node. Therefore, the service performed by the first node cannot be provided to the client.

If the operation response information is not received from the first node as a result of the determination of step 206, the normal node determines that the failure has occurred in the first node and causes another normal node to perform the service performed by the first node. (S210).

However, in the related art, a node in which an operating system is in a hang state is recognized as a normal node to other nodes.

In addition, since the ping command uses only two layers of the OSI layer, since the operating system is not affected by the row state, other nodes recognize the node as an operating state and thus do not serve the failed node.

Accordingly, an object of the present invention is to provide a node failure detection method and system that can quickly detect / overcome the nodes by checking whether the other node is disabled using the control board of the node in the system environment consisting of the heartbeat channel. It is.

According to an aspect of the present invention to achieve the above object, in the method for detecting a node failure of a system configured with a heartbeat channel, the resource management information for the node management module for managing the resource operation information of each node in the system When received by the management module, the management module analyzes the resource operation information to determine whether there is a resource whose operation has stopped, and if the resource whose operation has been stopped exists, determines that a failure occurs in the corresponding node. The method may further include providing a node failure detection method, wherein the management module transmits failure occurrence information of a failed node to another node and the other node performs a service performed by the failed node.

The failure occurrence information includes a unique node number of a failed node and a type of service performed by the failed node.

The service execution command includes a corresponding service registration command and then executes another node that receives the service execution command after registering the corresponding service according to the service registration command in the service execution command.

According to another aspect of the present invention, a system configured with a heartbeat channel, comprising: a node having a control board; A node failure detection system is provided that includes a management module connected to each node in the system to monitor the state of the node and perform a failure recovery command through the information received from the control board.

The control board includes an information module for collecting resource operation information of a node and a communication module for transmitting the collected resource operation information to the management module.

The management module is installed in the node or the user terminal, and analyzes resource operation information transmitted from the control board to determine whether the corresponding node has failed.

When the failure node is detected, the management module transmits failure occurrence information to another node to perform a service performed by the failed node in another node.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings.

3 is a diagram schematically illustrating a configuration of an information communication system for monitoring a node failure according to an embodiment of the present invention.

Referring to FIG. 3, an information communication system monitors a plurality of nodes 300a, 300b,..., 300n, and 300, a heartbeat network for monitoring the nodes 300 and exchanging information with each other. Shared disk 320 for sharing, the management module 310 for monitoring the status of the node 300 and performing a failover.

Each node 300 is connected through a heartbeat channel, and each node 300 has a built-in control board.

The control board includes an information module for collecting operation information of a CPU, a memory, and a hard disk, which are resources of a node, and a communication module for communicating with the management module 310. The information module collects operation information of the resource corresponding to the environment information preset by the user in the management module 310 and transmits the operation information to the management module 310 through the communication module.

The management module 310 analyzes the resource operation information transmitted from the control board, determines whether the corresponding node has failed, and informs another node about a node determined to have failed, so that another node provides services of the failed node. Let's do it.

The management module 310 is installed in a node or a user terminal and connected to the control board.

In addition, the management module 310 includes a management program including a resource selection menu for selecting a resource node to be monitored, and a service registration menu for performing a service being performed at a failed node in another node.

Accordingly, the user may change his or her preference information on the node resource by selecting the desired resource node in the management module 310.

4 is a flowchart illustrating a method of changing environment configuration information for node resources according to an embodiment of the present invention.

Referring to FIG. 4, the management module executes a management program to connect with a control board (S400).

Then, the management module authenticates the user (S402), and transmits a resource operation information request command of each node to the control board (S404). The user authentication receives a user's unique number, password, etc. from the user to determine whether the received information is registered in advance to perform user authentication.

Then, the control board transmits the resource operation information corresponding to the resource operation information request command, and the management module receives the resource operation information transmitted from the control board (S406). The resource operation information includes operation information of all resources included in the node.

The management module then displays the received node resource operation information (S408), and inquires whether the user wants to change the configuration information on the node resource (S410). That is, the management module inquires whether the user wants the operation information for all resources or the desired operation information for some resources.

As a result of the query in step 410, when the user selects and releases a resource in order to change environment setting information for the node resource, the management module selects and releases a resource to the node resource by selecting and releasing a resource performed by the user. The environment setting information is changed (S412).

Then, the control board transmits only resource operation information corresponding to the changed environment setting information to the management module.

5 is a flowchart illustrating a failure determination method of a node according to an exemplary embodiment of the present invention.

Referring to FIG. 5, when a management program is executed and connected to a control board (S500), the management module receives node resource operation information corresponding to preset configuration information from the control board (S502). That is, the management module transmits a node resource operation information request command to the control board after accessing the control board. Then, the control board collects resource operation information corresponding to the node resource operation information request command and transmits it to the management module.

After performing step 502, the management module analyzes the received node resource operation information to determine whether there is a resource whose operation is stopped (S504).

If there is a resource whose operation is stopped as a result of the determination in step 504, the management module determines that a failure has occurred in the corresponding node and transmits failure occurrence information to another node (S506). Here, the failure occurrence information includes a failure node, a type of service performed by the failure node, and the like.

After performing step 506, the management module transmits a command to perform a service performed by the failed node to another node (S508). At this time, the management module registers the corresponding service in order to allow another node to perform the service performed by the failed node. That is, the service execution command may include a corresponding service registration command.

The node receiving the service execution command registers and performs the corresponding service.

If there is no resource whose operation is stopped as a result of the determination of step 504, the management module determines that the node is normal (S510).

The present invention is not limited to the above embodiments, and many variations are possible by those skilled in the art within the spirit of the present invention.

 As described above, according to the present invention, it is possible to provide a node failure monitoring method and system that can provide continuous service by quickly detecting / breaking a node failure by using it in a high availability or clustering system because it is a failure detection of a system. .

In addition, according to the present invention, even when system resources are scarce, resource availability can be increased by increasing resources, and since the resources are checked by hardware signals, it is possible to provide a node failure monitoring method and system which is not expensive in implementation.

In addition, according to the present invention, since the operating system can know whether a failure occurs even if the normal or a specific service operates abnormally, it can take a quick action to provide a continuous service, the node that can increase the computational efficiency of the user Failure methods and systems can be provided.

Claims (9)

In the method for detecting a node failure of a system configured with a heartbeat channel, When the resource operation information for a node is received by the management module managing resource operation information of each node in the system, the management module analyzes the resource operation information to determine whether there is a resource whose operation is stopped. ; Determining that a failure occurs in a corresponding node when there is a resource whose operation is stopped as a result of the determination; And transmitting, by the management module, the failure occurrence information of the failed node to another node, and performing the service performed by the failed node. The method of claim 1, The failure occurrence information includes a node identification number of a failed node and a type of service performed by the failed node. The method of claim 1, The service failure command includes a corresponding service registration command. The method according to claim 1 or 3, And receiving the service execution command from another node in response to a service registration command in the service execution command and executing the corresponding service. In a system configured with a heartbeat channel, A node with a built-in control board; And a management module connected to each node in the system to monitor the state of the node through the information received from the control board and to perform a disaster recovery command. The method of claim 5, And the control board comprises an information module for collecting resource operation information of a node and a communication module for transmitting the collected resource operation information to the management module. The method of claim 5, The management module node failure detection system, characterized in that installed in the node or the user terminal. The method according to claim 5 or 7, The management module node failure detection system, characterized in that for determining the failure of the node by analyzing the resource operation information transmitted from the control board. The method of claim 8, The management module, when detecting a node in which a failure occurs, transmits failure information to another node to perform a service performed by the failed node in another node.

Images