Support Questions
Find answers, ask questions, and share your expertise
Announcements
Alert: Welcome to the Unified Cloudera Community. Former HCC members be sure to read and learn how to activate your account here.

How to recover from a failed NameNode move?

Solved Go to solution

How to recover from a failed NameNode move?

New Contributor

Ambari has a handy wizard for helping move a NameNode from one machine to another. In the event something were to go wrong with the move in, say, a NameNode HA enviornment are there any recommendations on how to recover and restore the NameNode setup back to its original server? Would the move NameNode wizard be the best approach to putting the server back or would the Ambari API be a better approach?

1 ACCEPTED SOLUTION

Accepted Solutions

Re: How to recover from a failed NameNode move?

As per a Support note:

"You can use the Move NameNode wizard in Ambari. This will move the NameNode but only according to Ambari.

After this has been successfully completed (with the NameNode down) then you should move all the files in the old namenode edits directory (dfs.namenode.name.dir) to the new NameNode in the directory configured. The permissions of these files will be hdfs:hadoop (by default) but the owner should be the user who runs your NameNode & the group will be the hadoop primary group.

After this is done, then the NameNode is ready to start."

The most important thing is to ensure that you have a backup of all the images and edits in dfs.namenode.name.dir. Then if anything happens you can you can revert back to that.

2 REPLIES 2

Re: How to recover from a failed NameNode move?

@Kent Baxley I think little bit more information will help. Is NN HA part of the picture?

Re: How to recover from a failed NameNode move?

As per a Support note:

"You can use the Move NameNode wizard in Ambari. This will move the NameNode but only according to Ambari.

After this has been successfully completed (with the NameNode down) then you should move all the files in the old namenode edits directory (dfs.namenode.name.dir) to the new NameNode in the directory configured. The permissions of these files will be hdfs:hadoop (by default) but the owner should be the user who runs your NameNode & the group will be the hadoop primary group.

After this is done, then the NameNode is ready to start."

The most important thing is to ensure that you have a backup of all the images and edits in dfs.namenode.name.dir. Then if anything happens you can you can revert back to that.

Don't have an account?
Coming from Hortonworks? Activate your account here