Single-point fault processing method, system and equipment and computer readable storage medium

Abstract

The invention discloses a single-point fault processing method. The method comprises the steps of acquiring all resource scheduling managers in a system; selecting one of all the resource scheduling managers as an activated state resource scheduling manager to process the request from the client, and setting the remaining resource scheduling managers as standby state resource scheduling managers;detecting whether the activated state resource scheduling manager is abnormal or not; and if so, selecting one of the standby state resource scheduling managers to take over the work of the activatedstate resource scheduling manager. The Hadoop high availability can be achieved without the need that each resource scheduling manager must be in an activated state, it is guaranteed that the resourcescheduling managers can still work after a single node of the server breaks down, and excessive occupation of resources is avoided. The invention further provides a single-point fault processing system and equipment and a computer readable storage medium, which have the above beneficial effects.

Description

technical field [0001] The present application relates to the field of single point fault handling, and in particular to a method, system, device and computer-readable storage medium for single point fault handling. Background technique [0002] Hadoop is a distributed system basic framework capable of processing large amounts of data, which is reliable, efficient and scalable. It mainly consists of Hadoop Distributed File System (Hadoop Distributed File System, HDFS) and MapReduce. Among them, MapReduce is a distributed computing framework, mainly used for parallel computing of large-scale data sets, which is mainly divided into Map phase (Map operation process) and Reduce phase (Reduce operation process). [0003] Hadoop can connect to distributed storage systems, integrate the distributed storage capabilities of distributed storage systems with the powerful parallel task processing capabilities of MapReduce in Hadoop, and combine HBASE column-oriented unstructured databa...

AI Technical Summary

Problems solved by technology

In order to ensure the normal and efficient operation of the Hadoop docking system, Hadoop achieves high availability by starting multiple Resource Manager (RM) processes, so that when one or more RMs are abnormal, the Hadoop MapReduce task can still proceed normally. However, This way of achieving high availability requires that each RM must be in an active state, resulting in excessive resource occupation

Images