The funny thing is....the figure as shown below... (The cluster is hadoop HA using HDP2) It was caused by a mis-operation...... At first, I wanna move a standby namenode to another host using Ambari (Click move nn in Service Actions), BUT I FORGOT RUNNING THE MANUAL COMMANDS!!! So I have 3 nns then........ and the cluster went down. By fixing my cluster, I moved the standby namenode back to the original host and restarted the cluster, and it worked!!!! Now my cluster is running normally and no data lost, but the Ambari dashboard still shows 3 nns. (I used backstage commands to check the states of my nns and it was good(nice)) 😄 I wanna fix the ambari service dashboard back to have only two nns, how should i do???? ( Or how can i delete disabled NameNode+ZKFC manully?) PLS HELP ME, MY BOSS IS ON FIRE SEEING THIS!!! TUT ... View more