HDFS Disk Balancer 介绍&使用

一、介绍

一句话介绍：用于 HDFS DataNode 单节点内多个数据盘数据均衡。
官网介绍如下：

Diskbalancer is a command line tool that distributes data evenly on all disks of a datanode. This tool is different from Balancer which takes care of cluster-wide data balancing. Data can have uneven spread between disks on a node due to several reasons. This can happen due to large amount of writes and deletes or due to a disk replacement. This tool operates against a given datanode and moves blocks from one disk to another. [1]

Diskbalancer 是一个命令行工具，用于在数据节点的所有磁盘上均匀分布数据。这个工具与 Balancer 不同，Balancer 负责整个集群的数据平衡。由于多种原因，数据在节点的磁盘之间可能分布不均匀。这可能是由于大量的写入和删除，或者由于磁盘替换引起的。该工具针对给定的数据节点操作，并将数据块从一个磁盘移动到另一个磁盘。

传送门： HDFS Diskbalancer 官方文档

二、配置

dfs.disk.balancer.