I have a 8 node hadoop cluster (CDH 5.2.0) of which i'm using all the nodes as datanodes.
I observed this rather peculiar behavior where no writes are happening on one of the data nodes and the disk activity is non existent for the past few weeks, although it appears as an active (commissioned) node in the cloudera manager window. When i look at the :/dfshealth.html page and click on live data nodes, this node is marked as decommissioned. I have tried to decommission the node in the cloudera manager and also recommission it through the same. The problem still remains, Is there a way to correct this and bring back the node to active status. Any help/suggestions is highly appreciated.
Thanks