Data reading method and device and related equipment

Abstract

The invention discloses a data reading method, which is applied to a cache mirror image pair. The cache mirror image pair comprises a first node and a second node which are mirror images of each other. The method comprises the following steps: when the first node receives a data reading request, determining the current state of the first node; when the own current state is a mirror image recombination state, judging whether request data corresponding to the data reading request hits own cache or not; if yes, reading the request data from the self cache; if not, reading the request data from the cache of the second node; and feeding back the request data to a request terminal. By applying the technical scheme provided by the invention, when the cache mirror image pair is recombined, the newly added node can immediately share the front-end service pressure, and meanwhile, the problem of data inconsistency among the nodes in the recombining process is avoided. The invention further discloses a data reading device, a cache mirror image system and a computer readable storage medium which all have the above beneficial effects.

Description

technical field [0001] The present application relates to the technical field of data storage, in particular to a data reading method, and also to a data reading device, a cache mirroring system, and a computer-readable storage medium. Background technique [0002] The high-end storage field has extremely high requirements for "fault tolerance, high performance, and scalability". For example, for a system with four nodes: it is required that the performance drop should not exceed 20% when any node fails; When three nodes fail, the service is required not to be interrupted. [0003] For example, in a storage system, node 1 and node 2 originally formed a mirror pair to maintain a double copy of the cache. When node 1 fails, node 2 will reorganize the mirror pair with node 3. At this time, node 2 stores the node 3 does not have cache data, therefore, the caches of node 2 and node 3 can reach agreement only after the "unique cache" on node 2 needs to be placed or copied to node...

AI Technical Summary

Problems solved by technology

[0004] However, the problem with the current cache reorganization mirror pair scheme is that the new node can only receive front-end requests after the newly added node reaches a consensus with the cache of the original node, otherwise the front-end host can only read it on the new node. The old data originally stored by the new node

However, in the time window of waiting for the two nodes to reach cache consistency, the service capability of the newly added node is actually sacrificed, so that all front-end business pressure is on a single node, which has a negative impact on business performance and data security.

Images