Data balancing method and equipment based on fragment cluster capacity expansion

Abstract

The invention aims to provide a data balancing method and equipment based on fragment cluster capacity expansion. Compared with the prior art, when a node receiving an external request sent by an external application is a first node generated after capacity expansion and the first node does not store data corresponding to the external request, the first node sends request information to all other nodes of a fragment cluster; the second node storing the data corresponding to the external request responds to the request information, so that the first node and the external application obtain migration data and the data corresponding to the external request; and after the first node acquires and stores the migration data, the first node sends migration completion information to the second node, so that the second node deletes the migration data. In this way, the data balancing process is decomposed into the external service process, extra consumption of resources of the fragment cluster is not needed, and data balancing can be achieved by providing services for the outside for multiple times.

Description

technical field [0001] The present application relates to the field of computer technology, and in particular to a data balancing technology based on expansion of fragmented clusters. Background technique [0002] A key link in the expansion of sharded clusters is to achieve data balance. Existing methods for achieving data balance can be roughly divided into two categories: [0003] 1. Suspend external services during the data balance period, concentrate resources to complete data balance as quickly as possible, and then resume external services. Such an implementation method is technically simpler and has higher data security, but for some services, the suspension of data services may cause major business accidents, so data balance cannot be achieved through this method; [0004] 2. Data services are still provided externally during the data balancing period, but this method causes the cluster to consume some resources for data balancing. During this period, the performan...

AI Technical Summary

Problems solved by technology

Such an implementation method is technically simpler and has higher data security, but for some services, the suspension of data services may cause major business accidents, so data balance cannot be achieved through this method;

[0004] 2. Data services are still provided externally during the data balancing period, but this method causes the cluster to consume part of the resources for data balancing. During this period, the overall external service performance of the cluster will decline accordingly

Images