Home | About | Sematext search-lucene.com search-hadoop.com
NEW: Monitor These Apps!
elasticsearch, apache solr, apache hbase, hadoop, redis, casssandra, amazon cloudwatch, mysql, memcached, apache kafka, apache zookeeper, apache storm, ubuntu, centOS, red hat, debian, puppet labs, java, senseiDB
 Search Hadoop and all its subprojects:

Switch to Plain View
HBase >> mail # user >> Re: Region Servers Issue


+
Adam Smieszny 2013-01-02, 13:52
+
Jeff Bean 2013-01-02, 18:42
+
Dalia Sobhy 2013-01-02, 21:27
Copy link to this message
-
Re: Region Servers Issue
Hello Dalia,

     You can turn the balancer on through the shell by using
"balancer_switch true". Alternatively, you can use "balancer()" provided by
the HBaseAdmin class.
HTH

Best Regards,
Tariq
+91-9741563634
https://mtariq.jux.com/
On Thu, Jan 3, 2013 at 2:57 AM, Dalia Sobhy <[EMAIL PROTECTED]>wrote:

> How to run hbase balancer given that I am using cloudera manager 4.1
>
> Sent from my iPad
>
> On Jan 2, 2013, at 9:13 PM, "Jeff Bean" <[EMAIL PROTECTED]> wrote:
>
> > Hi Dalia,
> >
> > If all the regions are still on the first server you don't get the
> benefit
> > of the second server. Try running the hbase balancer after adding the
> > second server. Then check to make sure each server is serving the same
> > number of regions.
> >
> > Jeff
> >
> > On Wed, Jan 2, 2013 at 5:52 AM, Adam Smieszny <[EMAIL PROTECTED]>
> wrote:
> >
> >> Redirecting to hbase-user
> >>
> >> Thanks,
> >> Adam
> >>
> >>
> >> On Tue, Jan 1, 2013 at 4:29 PM, Dalia Hassan <[EMAIL PROTECTED]
> >wrote:
> >>
> >>> Dear all,
> >>>
> >>> I am trying to test the performance of hbase by increasing number of
> >>> nodes.
> >>>
> >>> I am running a Java Program that uses Aggregate Implementation
> >>> Coprocessor i.e count the number of rows based on a specific filter.
> Then I
> >>> am testing by increasing the number of region servers will the time
> taken
> >>> to execute this program will less or not ? Suppose it will be.
> >>>
> >>> But unfortunately, it produces the same time.
> >>>
> >>> What I am doing actually is:
> >>>
> >>>   1. I have the number of rows to be 4 million.
> >>>   2. Then I ran the code, given that the cluster contains one region
> >>>   server , the time was approximately 16 seconds.
> >>>   3. After that I added another region server to the cluster i.e so now
> >>>   cluster has 2 region servers.
> >>>   4. By running the code, it took the same app 16 seconds.
> >>>
> >>>
> >>> So my question is what to do to feel the difference ? Also by adding a
> >>> second region server, will the 2 million records move to it ? If so,
> how
> >>> much time will it take ?
> >>>
> >>> Thanks,
> >>>
> >>> Waiting for your responses asap.
> >>
> >>
> >>
> >> --
> >> Adam Smieszny
> >> Cloudera | Systems Engineer | http://www.linkedin.com/in/adamsmieszny
> >> 917.830.4156
> >>
>
NEW: Monitor These Apps!
elasticsearch, apache solr, apache hbase, hadoop, redis, casssandra, amazon cloudwatch, mysql, memcached, apache kafka, apache zookeeper, apache storm, ubuntu, centOS, red hat, debian, puppet labs, java, senseiDB