HI All,
I want to remove the disk from Cloudera permanently. So far these are the steps that I should follow.
1. Remove the disk from HDFS configuration dfs.data.dir, dfs.datanode.data.dir
2. Remove the disk physically as these are not used anymore.
Now the question arises is that how the data blocks will be replicated if the data is removed from the DataNode Default Group like /data01.
what will happen if all the blocks according to RF=3 available on /data01 on different server and now the disk is removed? what is the possibility of the blocks available on same /data01 with RF=3 ?