Mailing-List: contact user-help@cassandra.apache.org; run by ezmlm
Precedence: bulk
Reply-To: user@cassandra.apache.org
Received-SPF: pass (athena.apache.org: domain of agundabattula@gmail.com
 designates 209.85.212.44 as permitted sender)
MIME-Version: 1.0
In-Reply-To: <1352439442.7645.YahooMailNeo@web160901.mail.bf1.yahoo.com>
References: <1352439442.7645.YahooMailNeo@web160901.mail.bf1.yahoo.com>
Date: Tue, 13 Nov 2012 10:52:39 +1100
Message-ID: 
 <CAJBMQQjVxuWyXO3BHX9D8CMx6ez5+NHe-SQB8tDgr0cDgTkSPg@mail.gmail.com>
Subject: Re: read request distribution
From: Ananth Gundabattula <agundabattula@gmail.com>
To: user@cassandra.apache.org, Wei Zhu <wz1975@yahoo.com>
Content-Type: multipart/alternative; boundary=047d7b67665020849e04ce54ff52

--047d7b67665020849e04ce54ff52
Content-Type: text/plain; charset=ISO-8859-1

Hi all,

On an unrelated observation of the below readings, it looks like all the 3
nodes own 100% of the data. This confuses me a bit. We have a 12 node
cluster with RF=3 but the effective ownership is shown as 8.33 % .

So here is my question. How is the ownership calculated : Is Replica factor
considered in the ownership calculation ? ( If yes , then 8.33 % ownership
of a cluster seems wrong to me . If not 100% ownership for a node cluster
seems wrong to me. Am I missing something in the calculation?

Regards,
Ananth

On Fri, Nov 9, 2012 at 4:37 PM, Wei Zhu <wz1975@yahoo.com> wrote:

> Hi All,
> I am doing a benchmark on a Cassandra. I have a three node cluster with
> RF=3. I generated 6M rows with sequence  number from 1 to 6m, so the rows
> should be evenly distributed among the three nodes disregarding the
> replicates.
> I am doing a benchmark with read only requests, I generate read request
> for randomly generated keys from 1 to 6M. Oddly, nodetool cfstats, reports
> that one node has only half the requests as the other one and the third
> node sits in the middle. So the ratio is like 2:3:4. The node with the most
> read requests actually has the smallest latency and the one with the least
> read requests reports the largest latency. The difference is pretty big,
> the fastest is almost double the slowest.
> All three nodes have the exactly the same hardware and the data size on
> each node are the same since the RF is three and all of them have the
> complete data. I am using Hector as client and the random read request are
> in millions. I can't think of a reasonable explanation.  Can someone please
> shed some lights?
>
> Thanks.
> -Wei
>

--047d7b67665020849e04ce54ff52
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable

<div>Hi all,<div><br></div><div>On an unrelated observation of the below re=
adings, it looks like all the 3 nodes own 100% of the data. This confuses m=
e a bit. We have a 12 node cluster with RF=3D3 but the effective ownership =
is shown as 8.33 % .=A0</div>
<div><br></div><div>So here is my question. How is the ownership calculated=
 : Is Replica factor considered in the ownership calculation ? ( If yes , t=
hen 8.33 % ownership of a cluster seems wrong to me . If not 100% ownership=
 for a node cluster seems wrong to me. Am I missing something in the calcul=
ation?=A0</div>
<div><br></div><div>Regards,</div><div>Ananth</div><br><div class=3D"gmail_=
quote">On Fri, Nov 9, 2012 at 4:37 PM, Wei Zhu <span dir=3D"ltr">&lt;<a hre=
f=3D"mailto:wz1975@yahoo.com" target=3D"_blank">wz1975@yahoo.com</a>&gt;</s=
pan> wrote:<br>
<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex"><div><div style=3D"font-size:12pt;font-famil=
y:arial,helvetica,sans-serif"><div>Hi All,</div><div>I am doing a benchmark=
 on a Cassandra. I have a three node cluster with RF=3D3. I generated 6M ro=
ws with sequence=A0 number from 1 to 6m, so the rows should be evenly distr=
ibuted among the three nodes disregarding the replicates. <br>
</div><div style=3D"font-style:normal;font-size:16px;background-color:trans=
parent;font-family:arial,helvetica,sans-serif">I am doing a benchmark with =
read only requests, I generate read request for randomly generated keys fro=
m 1 to 6M. Oddly, nodetool cfstats, reports that one node has only half the=
 requests as the other one and the third node sits in the middle. So the ra=
tio is like 2:3:4. The node with the most read requests actually has the sm=
allest latency and the one with the least read requests reports the largest=
 latency. The difference is pretty big,
 the fastest is almost double the slowest.<br></div><div style=3D"font-styl=
e:normal;font-size:16px;background-color:transparent;font-family:arial,helv=
etica,sans-serif">All three nodes have the exactly the same hardware and th=
e data size on each node are the same since the RF is three and all of them=
 have the complete data. I am using Hector as client and the random read re=
quest are in millions. I can&#39;t think of a reasonable explanation.=A0 Ca=
n someone please shed some lights?</div>
<div style=3D"font-style:normal;font-size:16px;background-color:transparent=
;font-family:arial,helvetica,sans-serif"><br></div><div style=3D"font-style=
:normal;font-size:16px;background-color:transparent;font-family:arial,helve=
tica,sans-serif">
Thanks.</div><span class=3D"HOEnZb"><font color=3D"#888888"><div style=3D"f=
ont-style:normal;font-size:16px;background-color:transparent;font-family:ar=
ial,helvetica,sans-serif">-Wei<br></div></font></span></div></div></blockqu=
ote>
</div><br></div>

--047d7b67665020849e04ce54ff52--