hadoop多数据中心灾备

【文件属性】：
文件名称：hadoop多数据中心灾备
文件大小：1.1MB
文件格式：PDF
更新时间：2021-08-15 07:59:20
大数据 灾备 The basis of this solution is to have one or more mirror Hadoop clusters which will be continuously updated with the data from the primary cluster in either a synchronous method or an asynchronous method by utilizing and improving the existing HA feature, data block replication and data block pipelining. In this solution, we support both synchronously writing and asynchronously replication across datacenters for both namespace and data block.