We have 3 nodes in our existing Datastax cluster. Sometime the ring status showing wrongly, after dse service restart on wrong nodes the issue got solved.
Correct :
nodetool status
Datacenter: SearchGraph
=======================
UN 10.10.1.56 1015.41 MiB 1 ? 936a1ac0-6d5e-4a94-8953-d5b5a2016b92 rack1
UN 10.10.1.46 961.43 MiB 1 ? 3f41dc2a-2672-47a1-90b5-a7c2bf17fb50 rack1
UN 10.10.1.36 1013.72 MiB 1 ? 0822145f-4225-4ad3-b2be-c995cc230830 rack1
Wrong :
Datacenter: DC1
===============
?N 10.10.1.46 ? 1 ? null r1
?N 10.10.1.36 ? 1 ? null r1
Datacenter: SearchGraph
=======================
UN 10.10.1.56 1005.33 MiB 1 ? null rack1
Configuration Details is same on all nodes :
cat /etc/dse/cassandra/cassandra.yaml | grep endpoint_snitch:
endpoint_snitch: GossipingPropertyFileSnitch
cat /etc/dse/cassandra/cassandra-rackdc.properties |grep -E 'dc=|rack='
dc=SearchGraph
rack=rack1
Issue got Resolved after restarting DSE nodes with saved_caches and hints folder content removed. I had tried to change the DC name in cassandra-rackdc.properties and revert it back earlier, may be issue caused by these.Thanks