question about speed up the hdfs balancer

hi,maillist: i want to balancer for my HDFS cluster ,but the default speed is 1M/s , i want to set the option dfs.balance.bandwidthPerSec to 20M/s ,surpose i issue balancer command on node A ,i do not know if i just set the option on node A's hdfs-site.xml ,or i need to set the option

Re: question about speed up the hdfs balancer

You have to push the conf to all the datanodes and restart the datanodes. On Mon, Nov 17, 2014 at 5:03 PM, ch huang justlo...@gmail.com wrote: hi,maillist: i want to balancer for my HDFS cluster ,but the default speed is 1M/s , i want to set the option dfs.balance.bandwidthPerSec

Re: question about speed up the hdfs balancer

If you set bandwidth on one of the nodes, it takes effect on the whole cluster. On Mon, Nov 17, 2014 at 1:56 AM, Gordon Wang gw...@pivotal.io wrote: You have to push the conf to all the datanodes and restart the datanodes. On Mon, Nov 17, 2014 at 5:03 PM, ch huang justlo...@gmail.com wrote:

CDH 4.7 upgrade java 6 to java 7

Hi, I am planning to upgrade Java version to Java 7. I am using ubuntu server Here is what I’m planning to do : 1. Stop all hadoop services in cloudera manager. 2. Update java version on all nodes (update-alternatives --set java /path/to/java). 3. Start all hadoop services in cloudera manager

Permissions issue with launching MR job from Oozie shell

I hope this question is relevant to the Hadoop mailing list. If it should be asked to the Oozie community, then I apologize in advance. I'm trying to schedule an HBase MapReduce bulkload preparation job with Oozie. I need to use either the Oozie shell action or Java action since the driver class

Showing INFO ipc.Client: Retrying connect to server once hadoop is upgraded to cdh5.2.0

Upgraded my Hadoop cluster (CDH) to cdh5.2.0 But once I run my Job with iteration, It is showing after 1 st iterative Job. 14/11/17 09:29:44 INFO ipc.Client: Retrying connect to server: /xx.xx.xx.xx:. Already tried 0 time(s); retry policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=3,

About HDFS-RAID

All I have found the HDFS-RAID on JIRA, but I can't find any patch on JIRA attachment. https://issues.apache.org/jira/browse/HDFS-503 is there anybody know how to add this patch to Hadoop 2.2.0 release version ? Or can the patch be used on this version? or I need found another one ? Thanks

RE: About HDFS-RAID

HDFS-RAID is not maintained anymore and has been removed from Hadoop. Now we are re-design Erasure coding in HDFS, refer to https://issues.apache.org/jira/browse/HDFS-7285 Regards, Yi Liu From: Vincent,Wei [mailto:weikun0...@gmail.com] Sent: Tuesday, November 18, 2014 1:55 PM To: