Sometimes, a node may be down and unable to be restarted. Common scenarios involve the host experiencing a hardware failure, or a cloud instance being prematurely destroyed. In the case of 192.168.0.102 being unrecoverable, a cluster status would look like this:
nodetool statusDatacenter: ClockworkAngels===========================Status=Up/Down|/ State=Normal/Leaving/Joining/Moving-- Address Load Tokens Owns Host ID RackUN 192.168.0.101 114.26 GB 24 100.0% 0edb5efa... R40DN 192.168.0.102 111.62 GB 24 100.0% 38782ca0... R40UN 192.168.0.103 101.81 GB 24 100.0% e45b2ee0... R40
As data cannot be streamed from a down node, nodetool decommission will not work. At this point, the best option for removing the node from the ...