The correct way to do this is to create a snapshot of the data in the second cluster/node using the snapshot API and then use the restore API to import it into the first node/cluster.
This will not work and could lead to you losing all the data. Do not try this.
You are running a very old version which is less resilient than more recent versions, so this MIGHT work. It could however fail as this is not something you are supposed to do nor supported, so could lead you to lose the data as well.
In summary I would recommend you going down the snapshot route.
Apache, Apache Lucene, Apache Hadoop, Hadoop, HDFS and the yellow elephant
logo are trademarks of the
Apache Software Foundation
in the United States and/or other countries.