Mailing-List: contact user-help@cassandra.apache.org; run by ezmlm
Precedence: bulk
Reply-To: user@cassandra.apache.org
Received-SPF: pass (nike.apache.org: domain of tbarbugli@gmail.com designates
 209.85.217.179 as permitted sender)
MIME-Version: 1.0
In-Reply-To: 
 <CABNXB2AAtQA4JRkHEGOhUBtGXvq5dKfCBagYR9TS0m_xkJBtbA@mail.gmail.com>
References: 
 <CAMwLmJYZSscQ1oYYtbV__rpruDw4cdujS4kXM2j5bYRj5qdEAw@mail.gmail.com>
	<CABNXB2AAtQA4JRkHEGOhUBtGXvq5dKfCBagYR9TS0m_xkJBtbA@mail.gmail.com>
Date: Mon, 21 Jul 2014 17:13:48 +0200
Message-ID: 
 <CAMwLmJYqqxf0mTi1r8sjQGF5p-SUidtLAxgFPMz-ZJPweR8brA@mail.gmail.com>
Subject: Re: estimated row count for a pk range
From: tommaso barbugli <tbarbugli@gmail.com>
To: "user@cassandra.apache.org" <user@cassandra.apache.org>
Content-Type: multipart/alternative; boundary=001a11c376dcd18e2f04feb58d73

--001a11c376dcd18e2f04feb58d73
Content-Type: text/plain; charset=UTF-8

thank you for the reply; I was hoping for something with a bit less
overhead than the first solution; the second is not really an option for me.

On Monday, 21 July 2014, DuyHai Doan <doanduyhai@gmail.com> wrote:

> 1) Use separate counter to count number of entries in each column family
> but it will require you to manage the counting manually
> 2) SELECT DISTINCT partitionKey FROM ....  Normally this query is
> optimized and is much faster than a SELECT *. However if you have a very
> big number of distinct partitions it can be slow
>
>
> On Sun, Jul 20, 2014 at 6:48 PM, tommaso barbugli <tbarbugli@gmail.com
> <javascript:_e(%7B%7D,'cvml','tbarbugli@gmail.com');>> wrote:
>
>> Hello,
>> Lately I collapsed several (around 1k) column families in a bunch (100)
>> of column families.
>> To keep data separated I have added an extra column (family) which is
>> part of the PK.
>>
>> While previous approach allowed me to always have a clear picture of
>> every column family's size; now I have no other option than select all the
>> rows and make some estimation to guess the overall size used by one of the
>> grouped data in this CFs.
>>
>> eg.
>> SELECT * FROM cf_shard1 WHERE family = '1';
>>
>> Of course this does not work really well when cf_shard1 has some data in
>> it; is there some way perhaps to get an estimated count for rows matching
>> this query?
>>
>> Thanks,
>> Tommaso
>>
>
>

-- 
sent from iphone (sorry for the typos)

--001a11c376dcd18e2f04feb58d73
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

thank you for the reply; I was hoping for something with a bit=C2=A0less ov=
erhead than the=C2=A0<span></span>first solution; the second is not really =
an option for me.<div><br>On Monday, 21 July 2014, DuyHai Doan &lt;<a href=
=3D"mailto:doanduyhai@gmail.com">doanduyhai@gmail.com</a>&gt; wrote:<br>
<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex"><div dir=3D"ltr"><div>1) Use separate counte=
r to count number of entries in each column family but it will require you =
to manage the counting manually<br>
</div>2) SELECT DISTINCT partitionKey FROM ....=C2=A0 Normally this query i=
s optimized and is much faster than a SELECT *. However if you have a very =
big number of distinct partitions it can be slow<br>
</div><div class=3D"gmail_extra"><br><br><div class=3D"gmail_quote">On Sun,=
 Jul 20, 2014 at 6:48 PM, tommaso barbugli <span dir=3D"ltr">&lt;<a href=3D=
"javascript:_e(%7B%7D,&#39;cvml&#39;,&#39;tbarbugli@gmail.com&#39;);" targe=
t=3D"_blank">tbarbugli@gmail.com</a>&gt;</span> wrote:<br>

<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex"><div dir=3D"ltr">Hello,<div>Lately I collaps=
ed several (around 1k) column families in a bunch (100) of column families.=
</div>

<div>To keep data separated I have added an extra column (family) which is =
part of the PK.</div><div>
<br></div><div>While previous approach allowed me to always have a clear pi=
cture of every column family&#39;s size; now I have no other option than se=
lect all the rows and make some estimation to guess the overall size used b=
y one of the grouped data in this CFs.</div>


<div><br></div><div>eg.</div><div>SELECT * FROM cf_shard1 WHERE family =3D =
&#39;1&#39;;</div><div><br></div><div>Of course this does not work really w=
ell when cf_shard1 has some data in it; is there some way perhaps to get an=
 estimated count for rows matching this query?</div>


<div><br></div><div>Thanks,</div><div>Tommaso</div></div>
</blockquote></div><br></div>
</blockquote></div><br><br>-- <br>sent from iphone (sorry for the typos)<br=
>

--001a11c376dcd18e2f04feb58d73--