3

I try to remove node from cassandra cluster v.2.2.4. The 'nodetool decommission' was run on removed node.

Now I have following:

nodetool status
Datacenter: dc1
===============
Status=Up/Down
|/ State=Normal/Leaving/Joining/Moving
--  Address        Load       Tokens       Owns    Host ID                               Rack
UN  10.0.0.0.158  30.83 GB   256          ?       8b49b920-30c6-4b99-b00e-f1667e12386a  r1
UN  10.0.0.0.77   31.01 GB   256          ?       4c207524-8e8a-44a5-a428-ce5027406eb2  r1
UN  10.0.0.0.154  28.5 GB    256          ?       12ed1ee7-9425-4107-95e1-420067ef536d  r1
UN  10.0.0.0.152  28.69 GB   256          ?       7cd58f7f-ba4c-455c-80a8-a8dca856ac38  r1
UN  10.0.0.0.150  28.22 GB   256          ?       f59e8f67-1459-43ae-8469-7ac455b1d858  r1
UN  10.0.0.0.85   59.03 GB   256          ?       c372b9e6-922b-4198-aa60-558ae2a181bd  r1
UN  10.0.0.0.133  33.44 GB   256          ?       54fcb413-3a46-46ae-905d-53fb8c8e0534  r1

nodetool describecluster
Cluster Information:
        Name: testCluster1
        Snitch: org.apache.cassandra.locator.DynamicEndpointSnitch
        Partitioner: org.apache.cassandra.dht.Murmur3Partitioner
        Schema versions:
                15c69d7a-ff7b-3f3a-8190-a4632cfebe6b: [10.0.0.0.158, 10.0.0.0.77, 10.0.0.0.154, 10.0.0.0.152, 10.0.0.0.150, 10.0.0.0.133, 10.0.0.0.85]

                UNREACHABLE: [10.0.0.0.212]

Why is the removed node 10.0.0.0.212 marked 'UNREACHABLE'? How can this issue be resolved?

Dimaf
  • 653
  • 1
  • 11
  • 25
  • Looks like `nodetool decommission` didn't clean-up after itself. I answered a similar question a couple days ago. The same solution may help you as well: http://stackoverflow.com/questions/35751921/cassandra-host-in-cluster-with-null-id/35756401#35756401 – Aaron Mar 04 '16 at 17:15
  • 2
    `nodetool assassinate 10.0.0.0.212` - works. @Aoron, thank you. – Dimaf Mar 04 '16 at 17:23

0 Answers0