This article will introduce disaster recovery of data and dervices in E-MapReduce clusters


HDFS stores the data of each file in blocks, with each block holding multiple copies (three by default). HDFS also makes sure that these copies are stored in different frameworks. In most situations, HDFS stores the first copy in the local framework, the second in the same framework as the first but in different nodes, and the last copy in a different framework.

HDFS scans the data copies regularly. If it finds that a data copy has been lost, HDFS makes another to make sure the number of copies is stable. If a node that stores a copy has been lost, HDFS makes another node to recover the data in that node. In Alibaba Cloud, if you use cloud disks, each cloud disk has three data copies in the back-end. If any of them has an issue, the copies exchange and recover data to ensure reliability.

HDFS is a highly reliable file storage system that can store massive amounts of data. Based on the features of Alibaba Cloud, HDFS can also make backups of the data stored in OSS, providing even greater data reliability.


The core components of HDFS guarantee high availability by making sure that there are at least two nodes to back each other up, such as YARN, HDFS, HiveServer, or Hive Meta. In this way, whenever a node experiences an issue, the nodes can exchange and recover data to ensure that services are not impacted.