What are roles and responsibilities of Backup Node in Apache Hadoop?
What is Backup Node and how it works in Hadoop?
Backup node isn't a concept.
Several components have High-Availability setup, for example your other question about the NameNode.
ResourceManager, HiveServer, HBase Masters, and other components have availability considerations.
A backup node could be a separate node other than your cluster nodes, where you are keeping backup of you back-end databases. Such databases are required to run hadoop services such as Hive(metastore), Hue, Oozie, Sqoop. all these services require databases for storing table/partition/files information, scheduled jobs, or user database. These are some essential and important databases which currently do not participate in cluster HA setup or are not a part of it because such databases have their own replication or backup mechanism.