Failed to start secondary namenode

[#1]

I have done the configuration HA for YARN, I followed the wisard in enable ResourceManager HA to enable the nameode and Snamenode. and for some raison I cancel the wizard in the last step, I thought I can done it again later.

So when I cancel the config process all services turn off and I got lot of alerts (scary!! right).

so I tried to restart all the services but I could not. there is something wrong With my Snamenode.

in addition when I try to start the service I see that the service Snamenode in my standby namenode is in maintenance mode

This is the erreur msg:

2017-11-17 17:19:10,364 INFO  namenode.SecondaryNameNode (LogAdapter.java:info(45)) - registered UNIX signal handlers for [TERM, HUP, INT]
2017-11-17 17:19:10,770 FATAL namenode.SecondaryNameNode (SecondaryNameNode.java:main(673)) - Failed to start secondary namenode
java.io.IOException: Cannot use SecondaryNameNode in an HA cluster. The Standby Namenode will perform checkpointing.
	at org.apache.hadoop.hdfs.server.namenode.SecondaryNameNode.<init>(SecondaryNameNode.java:187)
	at org.apache.hadoop.hdfs.server.namenode.SecondaryNameNode.main(SecondaryNameNode.java:671)
2017-11-17 17:19:10,773 INFO  util.ExitUtil (ExitUtil.java:terminate(124)) - Exiting with status 1
2017-11-17 17:19:10,776 INFO  namenode.SecondaryNameNode (LogAdapter.java:info(45)) - SHUTDOWN_MSG: 

and this is the entire error msg:

error-body.txt

[#2]

Hi Sihi,

I also encountered similar issue while enabling HA on name-node. We fall into this issue usually because we miss some instructions while enabling HA , like in my case I missed one step to create checkpoint on second NN. 

To overcome this thing , please try to start the HA enabling process once again from Ambari UI . And follow every step again correctly. It will again try to create secondary NN and you will be able to bring back the cluster to green state. 

Thanks

Arun