Mailing-List: contact user-help@hbase.apache.org; run by ezmlm
Precedence: bulk
Reply-To: user@hbase.apache.org
Received-SPF: pass (athena.apache.org: domain of yuzhihong@gmail.com
 designates 74.125.82.42 as permitted sender)
MIME-Version: 1.0
In-Reply-To: 
 <CAE24rAdJdDOBY5RYvhr8NSVDkCPvGpfd=vmNGWGAgJ44-iQ-wA@mail.gmail.com>
References: 
 <CAE24rAdJdDOBY5RYvhr8NSVDkCPvGpfd=vmNGWGAgJ44-iQ-wA@mail.gmail.com>
Date: Fri, 25 Jan 2013 16:14:02 -0800
Message-ID: 
 <CALte62zA717TNLWYELd=aJa6WF-cfbyfKkfWkDhsprcRHYmmuQ@mail.gmail.com>
Subject: Re: Rule of thumb: Size of data to send per RPC in a scan
From: Ted Yu <yuzhihong@gmail.com>
To: user@hbase.apache.org
Content-Type: multipart/alternative; boundary=bcaec53d5aa3db640e04d425eb9e

--bcaec53d5aa3db640e04d425eb9e
Content-Type: text/plain; charset=ISO-8859-1

Looks like HBASE-2214 'Do HBASE-1996 -- setting size to return in scan
rather than count of rows -- properly' may help you.
But that is only in 0.96

Lars H presented some performance numbers in:
  HBASE-7008 Set scanner caching to a better default, disable Nagles
where default for "hbase.client.scanner.caching" changed to 100

Cheers

On Fri, Jan 25, 2013 at 3:59 PM, David Koch <ogdude@googlemail.com> wrote:

> Hello,
>
> Is there a rule to determine the best batch/caching combination for
> maximizing scan performance as a function of KV size and (average) number
> of columns per row key?
>
> I have 0.5kb per value (constant), an average of 10 values per row key -
> heavy tailed so some outliers have 100k KVs, around 100million rows in the
> table. The cluster consists of 30 region servers, 24gb of RAM each, nodes
> are connecting with a 1gbit connection. I am running Map/Reduce jobs on the
> table, also with 30 task trackers.
>
> I tried:
> cache: 1, no batching -> 14min
> cache 1000, batch 50 -> 11min
> cache 5000, batch 25 -> crash (timeouts)
> cache 2000, batch 25 -> 15min
>
> Job time can vary quite significantly according to whatever activity
> (compactions?) are going on in the background. Also, I cannot probe for the
> best combination indefinitely since there actual production jobs queued. I
> did expect a larger speed-up with respect to no caching/batching at all -
> is this unjustified?
>
> In short, I am looking for some tips for making scans in a Map/Reduce
> context faster :-)
>
> Thank you,
>
> /David
>

--bcaec53d5aa3db640e04d425eb9e--