Device access server cluster method, system, device and storage medium

Abstract

The invention provides a method, system, device and storage medium for a device to access a server cluster. The method includes the following steps: assigning hash ring partitions to each server in the cluster to obtain a hash ring partition table; obtaining the device to be connected Corresponding hash value; determine the server identification information corresponding to the device to be accessed; query the access information of the server, and return to the device to be accessed. The present invention realizes the automatic allocation of equipment and access servers, without the need for further allocation by the central node; when offline events such as server downtime or new servers occur, the mapping relationship between the hash ring partition and the server is automatically updated, and the server according to the update After the hash ring partition table automatically disconnects the connection with the device to be transferred, the device can re-query the access information of the server and re-connect, thus forming a device connection with no single point of failure, fast positioning and automatic recovery. Enter the server cluster solution.

Description

technical field [0001] The present invention relates to the technical field of server clusters, and in particular to a method, system, device and storage medium for connecting a device to a server cluster. Background technique [0002] In the monitoring system, with the continuous expansion of the front-end monitoring cameras, the front-end equipment access system is under great performance pressure. There are two traditional solutions: [0003] (1) if figure 1 As shown, the performance of the access server is enhanced, and better hardware servers are used to complete the front-end device access; [0004] (2) if figure 2 As shown in the figure, a hierarchical structure is used for device access, and the pressure is distributed to the device access system of the sub-nodes, and each front-end device is fixedly connected to the corresponding device server. [0005] The disadvantage of the first solution is that when the number of devices continues to increase and reaches t...

AI Technical Summary

Problems solved by technology

[0005] In the first solution, the disadvantage is that when the amount of equipment continues to increase and reaches the million level, the performance of the enhanced central server is still insufficient, and the cost performance of the central server is greatly reduced.

[0006] In the second solution, the defect is that there will be a single point of failure. When the device access system of a certain child node stops serving, the device cannot automatically connect to other access systems, and can only be located and specified from the central node according to the reservation information. The access service node of the device, and then send the corresponding information to the device

At the same time, the central node is also a large single point of failure. If there is a problem with the central node, the entire system cannot operate

[0007] There is an improved method for the second solution. When the scheduled access server of the device goes down, the device applies to the central server to connect to the new server node. However, this will cause the problem of device positioning. When the device After the connected server is changed, the central access node must maintain a dynamic change information from the device to the access server. When locating the device, it must be searched before it can communicate with the device, increasing the load on the central access node.

At the same time, the central access node needs to allocate access server nodes to unconnected devices according to the load of each access server, which increases business complexity

Images