Storage method, primary node, secondary nodes, and system comprising primary node and secondary nodes

Abstract

The embodiment of the invention relates to the technical field of distributed storage, and discloses a storage method, a primary node, secondary nodes and a system comprising the primary node and thesecondary nodes. The storage method comprises the following steps: selecting a standby node from secondary nodes of a node group, wherein the node group is a node group where the main node is located,and the standby node is used for replacing the main node when the main node is offline; monitoring whether the number of the online nodes of the node group is smaller than a first threshold value ornot, and if the number of the online nodes of the node group is smaller than the first threshold value, selecting a target merging node group from the system to merge the node group and the target merging node group into a new node group, wherein the combined new node group stores the data stored in each node group before combination; and monitoring whether the number of the online nodes of the node group is greater than a second threshold, and if the number is greater than the second threshold, splitting the node group. According to the embodiment of the invention, the availability of the decentralized distributed storage system can be further improved.

Description

technical field [0001] The present invention relates to the technical field of distributed storage, in particular to a storage method, primary and secondary nodes and a system including the same. Background technique [0002] The distributed storage system is one of the cornerstone systems of the Internet. At present, well-known cloud vendors and large Internet companies have their own distributed storage systems. Based on the distributed storage system, it can provide network disk services to ordinary users, which can provide larger storage capacity and higher fault tolerance. However, existing services of this type are based on centralized distributed storage systems. Since the correct execution of centralized distributed storage systems (such as security, availability, etc.) depends on a certain central organization / module, once the cloud If there is a problem with the manufacturer's hardware or network, users will not be able to access the data stored in the centralize...

AI Technical Summary

Problems solved by technology

[0006] The inventor found that the distributed hash table has at least the following problems: in the case of decentralization, the method itself does not guarantee that the corresponding hash table can be found

The distributed hash table itself does not ensure that the hash table entries that the target needs to query can be found through routing. In other words, this method cannot ensure that the online nodes in the network keep a complete hash table

Therefore, this kind of decentralized distributed storage system still does not have high availability

Images