hadoop-mapreduce-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Michael Namaiandeh <mnamaian...@healthcit.com>
Subject RE: Two Datanodes - Incompatible Cluster IDs
Date Tue, 11 Jun 2013 15:55:53 GMT
Brilliant, Ian! That worked. Now, all my nodes are showing up as live.



From: Ian Wrigley [mailto:ian@cloudera.com]
Sent: Tuesday, June 11, 2013 11:43 AM
To: user@hadoop.apache.org
Subject: Re: Two Datanodes - Incompatible Cluster IDs

If you just blow away the contents of the dfs.data.dir directories then start the DataNode
daemon again you should be fine. (That's where the ClusterID is stored. When they next connect
to the NameNode, because they don't have one now that you've deleted it, they'll be given
the one from the NameNode.)

Regards

Ian.

On Jun 11, 2013, at 10:38 AM, Michael Namaiandeh <mnamaiandeh@healthcit.com<mailto:mnamaiandeh@healthcit.com>>
wrote:


I am trying to setup a 4 node Cloudera Hadoop cluster. However; two of my data nodes are showing
up as "dead" nodes. After looking at the logs, I found that the dead nodes have a different
ClusterID than my working/"live" nodes.

How do I configure the dead nodes with the correct ClusterID.  I can format the dead nodes,
if possible. Any help would be greatly appreciated. Thanks.

Mike



---
Ian Wrigley
Sr. Curriculum Manager
Cloudera, Inc
Cell: (323) 819 4075


Mime
View raw message