CN102722146B - Distributed system control structure with failure protection function, and failure protection method - Google Patents

Abstract

The invention provides a distributed system control structure with failure protection and a failure protection method. The failure protection setting of the original connection in the distributed system is set. Starting from the second layer of the distributed system, the adjacent nodes of the same layer are connected. The connected lower-level nodes detect whether the communication or management is invalid according to the information returned by the lower-level nodes; if the communication or management is found to be invalid, the adjacent lower-level nodes are controlled to restore the failed communication or management. The present invention is suitable for occasions requiring high safety and reliability, such as fire alarm systems, mine safety systems and the like.

Description

Distributed System Control Structure with Failsafe and Failsafe Method

technical field

The invention relates to the field of distributed system control, and also relates to a distributed system control failure protection method.

Background technique

The application of distributed system control is very extensive, so the reliability and security of distributed system control are particularly important. At present, the reliability of distributed system control mainly comes from the fault-tolerant mechanism of its structure itself, that is, if the global communication or management is interrupted, the local station can still maintain work, but it cannot resume the interrupted communication or management. Therefore, an effective protection method is needed, so that the communication or control controlled by the distributed system can be effectively protected after failure, even if the communication and control are still valid.

The existing distributed system control structure is shown in Figure 1. The reliability of the structure for communication or management depends on the structure of the distributed system control itself, and when the system is interrupted in communication or management, it cannot be recovered. Therefore, a method is needed to solve the above problems.

Contents of the invention

The object of the present invention is to provide a distributed system control structure with high reliability and safety and fail-safe. The purpose of the present invention is also to provide a failure protection method for distributed system control.

The purpose of the present invention is achieved like this:

The control structure of the distributed system with failure protection of the present invention is: the adjacent nodes of the second layer of the distributed system are connected sequentially; among the nodes of the third layer of the distributed system, all nodes belonging to the same node on the upper layer are controlled Nodes are connected sequentially; the other layers of the distributed system are connected in the same way as the third layer.

The failure protection method of the distributed system control of the present invention comprises:

Adjacent nodes on the second layer of the distributed system are connected sequentially; among nodes on the third layer of the distributed system, nodes under the control of the same node belonging to the upper layer are connected in sequence; the connection mode of other layers of the distributed system is the same as that of the first Three layers are the same;

When the upper layer node controls a certain node in the lower layer, it will send information to all nodes controlled by the lower layer at the same time; if it is the controlled node, it will return a confirmation message after executing the control task, and if it is not the controlled node, it will also return a confirmation message;

When it is detected that a lower node does not return a confirmation message, that is, the communication or management failure is judged to be a failure node, the node adjacent to the failure node controls the failure node, and sends the communication or management signal sent by the upper node to the failure node. Restore the failed communication or management; after the upper node receives the confirmation message sent by the failed node again, it jumps back to the original communication connection.

The present invention sets the failure protection for the original connection in the distributed system. Starting from the second layer of the distributed system, the adjacent nodes of the same layer are connected and set; The lower-level nodes connected to it detect whether the communication or management is invalid according to the information returned by the lower-level nodes; if the communication or management failure is found, the adjacent lower-level nodes are controlled to restore the failed communication or management.

The invention can increase the reliability and security of the distributed system control through the failure protection of the distributed system nodes. When the distributed system control fails, it can quickly restore communication or control, which is especially suitable for occasions with high safety and reliability requirements, such as fire alarm systems and mine safety systems.

Description of drawings

Fig. 1 is a schematic diagram of the existing distributed system structure.

Fig. 2 is a schematic diagram of the failure protection of the distributed system of the present invention.

Detailed ways

Below in conjunction with accompanying drawing, the present invention will be further described:

Combined with Figure 2, the control structure of the distributed system with fail-safe is divided into 3 layers: the first layer has 1 master, the second layer has n slaves, and the third layer has m under the control of each slave. a control unit. For the second layer, slave 1 to slave n are sequentially connected, and in the third layer, the m control units under the control of each slave are connected sequentially, and the control units under the control of different control slaves are not connected.

Detect communication or management failures. When the upper node sends a control command to a node in the lower layer, the command will be sent to all lower nodes controlled by the upper layer node. After the lower node receives the command, if it is a control command for itself, it will execute the control command and return a confirmation message; if it is not a control command for itself, it will still return a confirmation message. The upper layer node judges whether the communication is valid according to the returned confirmation message. If no confirmation message is received, the communication is considered invalid, otherwise the communication is considered valid.

Restoration of failed communications or management. If the upper layer node finds that there is a confirmation message that has not been returned, it will consider the communication to be invalid, and the failsafe will be enabled at this time. Since nodes at the same layer are connected sequentially, it is necessary to select a failsafe channel when a failure occurs. The selection method is as follows: if node i fails, judge whether node i is the last node of the current layer node, if it is the last node, the i-1th node will send the control information to node i to restore the failure; if i is not the last node node, the i+1th node sends the control information to the i node. When the upper node finds that the failed node has returned to normal, the original channel will be re-enabled, and the adjacent lower nodes will no longer protect the original failed node.

Claims (1)

1. A fail-safe method for distributed system control, characterized in that: Adjacent nodes on the second layer of the distributed system are connected sequentially; among nodes on the third layer of the distributed system, nodes under the control of the same node belonging to the upper layer are connected in sequence; the connection mode of other layers of the distributed system is the same as that of the first Three layers are the same; When the upper layer node controls a certain node in the lower layer, it will send information to all nodes controlled by the lower layer at the same time; if it is the controlled node, it will return a confirmation message after executing the control task, and if it is not the controlled node, it will also return a confirmation message; When it is detected that a lower node does not return a confirmation message, that is, the communication or management failure is judged to be a failure node, the node adjacent to the failure node controls the failure node, and sends the communication or management signal sent by the upper node to the failure node. Restore the failed communication or management; after the upper node receives the confirmation message sent by the failed node again, it jumps back to the original communication connection.

Images