Not sure if this is normal or not, but I see kylin metadata is on a single region server (DN & RS on node).
if this datanode goes down... it appears kylin isn't able to pull jobs for monitor or complete jobs? hbase requests: kylin_metadata,,1467291011009.2cc83fc3fb51700a8a9884e5c5401e20. 5534555 49271 ^^
