Home | About | Sematext search-lucene.com search-hadoop.com
 Search Hadoop and all its subprojects:

Switch to Threaded View
HBase, mail # user - Hbase Read/Write throughput measure


Copy link to this message
-
RE: Hbase Read/Write throughput measure
Dalia Sobhy 2013-02-02, 14:47

any help ??!!!!!!

> Date: Fri, 1 Feb 2013 17:15:45 -0500
> Subject: Re: Hbase Read/Write throughput measure
> From: [EMAIL PROTECTED]
> To: [EMAIL PROTECTED]
>
> When you will have figure how the command is working, you will have to
> understand the output ;)
>
> This is how I just tried it from my HBase directory:
> bin/hbase org.apache.hadoop.hbase.PerformanceEvaluation randomRead 10
>
> And this is the output (last lines):
> 10066326 56314
> 10171183 56385
> 10276040 56579
> 10380897 56553
>
> Does ask me what it means ;) I have, so far, no idea.
>
> JM
>
> 2013/2/1, Dalia Sobhy <[EMAIL PROTECTED]>:
> > Dear Mohamed,
> >
> > I checked this link:
> > http://wiki.apache.org/hadoop/Hbase/PerformanceEvaluation
> >
> > But how to use the commands there is no clue, note that I am using cloudera
> > manager.
> >
> > The other yahoo link I couldn't understand how to use it.
> >
> > So any help please??
> >
> > Sent from my iPad
> >
> > On Feb 1, 2013, at 2:32 PM, "Mohammad Tariq" <[EMAIL PROTECTED]> wrote:
> >
> >> Hello Dalia,
> >>
> >>          I think the easiest way to measure the read/write throughput is
> >> to use "PerformanceEvaluation" tool that comes with the Hbase
> >> distribution. It spawns a map-reduce job to do the reads/writes in
> >> parallel. Apart from this there are several other ways to benchmark your
> >> Hbase cluster like YSCB <https://github.com/brianfrankcooper/YCSB>. You
> >> might find this
> >> link<http://wiki.apache.org/hadoop/Hbase/PerformanceEvaluation> useful
> >> that talks about Hbase performance.
> >>
> >> HTH
> >>
> >> Warm Regards,
> >> Tariq
> >> https://mtariq.jux.com/
> >> cloudfront.blogspot.com
> >>
> >>
> >> On Fri, Feb 1, 2013 at 5:41 PM, Dalia Sobhy
> >> <[EMAIL PROTECTED]>wrote:
> >>
> >>>
> >>> Dear all,
> >>>
> >>> I want to measure the read/write throughput for a code on a cluster of
> >>> 10
> >>> nodes. So is there any code or way to measure it?
> >>>
> >>> I have seen in a cloudera-based presentation that hbase read/write
> >>> throughput = millions queries per second.
> >>>
> >>> So any help please??
> >>>
> >>> Thanks
> >>>
> >>>
> >>> Best Regards,
> >>> Dalia Sobhy
> >>>
> >