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 chen.daqi@gmail.com designates
 74.125.82.44 as permitted sender)
DomainKey-Signature: a=rsa-sha1; c=nofws;
        d=gmail.com; s=gamma;
        h=mime-version:in-reply-to:references:date:message-id:subject:from:to
         :content-type;
        b=FzWl8rl/UMFhxTTO9rFuC2CQbHYdkDanobF5UgrKeWuJyeDZWG5srIvZ3NqFIEADwP
         a2zY5HD9nohFTJedgPF+jJB34xOyUPyjX1Rk6x+FHekfn6ZRlRn9Z4buIVyHtCeuIzOI
         vnCEj7yc3QPz386rmZRR/QRZH+IOzGxEI3jAU=
MIME-Version: 1.0
In-Reply-To: <AANLkTikfzssQc92dtYbBbnPOQG3tmUNa0=imjWbnunmU@mail.gmail.com>
References: <AANLkTinsXh=CHN-O4DLXd+-jFNoxfXkPBXKi=FC0Ro82@mail.gmail.com>
	<AANLkTikfzssQc92dtYbBbnPOQG3tmUNa0=imjWbnunmU@mail.gmail.com>
Date: Wed, 18 Aug 2010 11:52:07 +0800
Message-ID: <AANLkTinm4XeJM1_3c_VDj6EZ-a7Dd0o0CT+G+A6ZJZEc@mail.gmail.com>
Subject: Re: cassandra for a inbox search with high reading qps
From: Chen Xinli <chen.daqi@gmail.com>
To: user@cassandra.apache.org
Content-Type: multipart/alternative; boundary=0016e65b5e3457e0c8048e10fd88

--0016e65b5e3457e0c8048e10fd88
Content-Type: text/plain; charset=ISO-8859-1

I'm using cassandra 0.6.4; there's a configuration option
DoConsistencyChecksBoolean in storage-conf.xml.
Is't that for read-repair ?

I will do a test for WRITE QUORUM, READ.ONE if it can meet our requirements.

2010/8/18 Edward Capriolo <edlinuxguru@gmail.com>

> On Tue, Aug 17, 2010 at 10:55 PM, Chen Xinli <chen.daqi@gmail.com> wrote:
> > Hi,
> >
> > We are going to use cassandra for searching purpose like inbox search.
> > The reading qps is very high, we'd like to use ConsitencyLevel.One for
> > reading and disable read-repair at the same time.
> >
> > For reading consistency in this condition, the writing should use
> > ConsistencyLevel.ALL. But the writing will fail if one node fails.
> > We want such a ConsistencyLevel for writing/reading that :
> > 1. writing will success if there is node alive for this key
> > 2. reading will not forward to a node that's just recovered and doing
> hinted
> > handoff
> >
> > So that, if some node fails, others nodes for replica will receive the
> data
> > and surve reading successfully;
> > when the failure node recovers,  it will receive hinted handoff from
> other
> > nodes and it'll not surve reading until hinted handoff is done.
> >
> > Does cassandra support the cases already? or should I modify the code to
> > meet our requirements?
> >
> > Thanks for any advices!
> >
> > --
> > Best Regards,
> > Chen Xinli
> >
>
> >>The reading qps is very high, we'd like to use ConsitencyLevel.One for
> reading and disable read-repair at the same time.
> You can not disable read repair, all reads no matter the
> ConsistencyLevel always repair. The CL only controls how many nodes to
> read from before returning data to the client.
>
> These two statements contradict.
> > For reading consistency in this condition, the writing should use
> ConsistencyLevel.ALL. But the writing will fail if one node fails.
> > 1. writing will success if there is node alive for this key
>
> Also regardless of the write ConsistencyLevel all writes are written
> to all nodes. If a target node is down HintedHandoff will queue the
> write up for when the node restarts * in (6.3 you can turn off Hinted
> Handoff)
>
> You may want to WRITE QUORUM, READ.ONE, with RF=3 you would need two
> failed nodes before seeing an UnavailableException. You still have
> pretty strong consistency (depending how you look at it) and fast
> reads.
>
> Check the IRC logs Ben "schooled" me over this a couple of days ago.
>


-- 
Best Regards,
Chen Xinli

--0016e65b5e3457e0c8048e10fd88
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable

I&#39;m using cassandra 0.6.4; there&#39;s a configuration option DoConsist=
encyChecksBoolean in storage-conf.xml.<br>Is&#39;t that for read-repair ?<b=
r><br>I will do a test for WRITE QUORUM, READ.ONE if it can meet our requir=
ements.<br>
<br><div class=3D"gmail_quote">2010/8/18 Edward Capriolo <span dir=3D"ltr">=
&lt;<a href=3D"mailto:edlinuxguru@gmail.com">edlinuxguru@gmail.com</a>&gt;<=
/span><br><blockquote class=3D"gmail_quote" style=3D"margin: 0pt 0pt 0pt 0.=
8ex; border-left: 1px solid rgb(204, 204, 204); padding-left: 1ex;">
<div><div></div><div class=3D"h5">On Tue, Aug 17, 2010 at 10:55 PM, Chen Xi=
nli &lt;<a href=3D"mailto:chen.daqi@gmail.com">chen.daqi@gmail.com</a>&gt; =
wrote:<br>
&gt; Hi,<br>
&gt;<br>
&gt; We are going to use cassandra for searching purpose like inbox search.=
<br>
&gt; The reading qps is very high, we&#39;d like to use ConsitencyLevel.One=
 for<br>
&gt; reading and disable read-repair at the same time.<br>
&gt;<br>
&gt; For reading consistency in this condition, the writing should use<br>
&gt; ConsistencyLevel.ALL. But the writing will fail if one node fails.<br>
&gt; We want such a ConsistencyLevel for writing/reading that :<br>
&gt; 1. writing will success if there is node alive for this key<br>
&gt; 2. reading will not forward to a node that&#39;s just recovered and do=
ing hinted<br>
&gt; handoff<br>
&gt;<br>
&gt; So that, if some node fails, others nodes for replica will receive the=
 data<br>
&gt; and surve reading successfully;<br>
&gt; when the failure node recovers,=A0 it will receive hinted handoff from=
 other<br>
&gt; nodes and it&#39;ll not surve reading until hinted handoff is done.<br=
>
&gt;<br>
&gt; Does cassandra support the cases already? or should I modify the code =
to<br>
&gt; meet our requirements?<br>
&gt;<br>
&gt; Thanks for any advices!<br>
&gt;<br>
&gt; --<br>
&gt; Best Regards,<br>
&gt; Chen Xinli<br>
&gt;<br>
<br>
&gt;&gt;The reading qps is very high, we&#39;d like to use ConsitencyLevel.=
One for reading and disable read-repair at the same time.<br>
</div></div>You can not disable read repair, all reads no matter the<br>
ConsistencyLevel always repair. The CL only controls how many nodes to<br>
read from before returning data to the client.<br>
<br>
These two statements contradict.<br>
<div class=3D"im">&gt; For reading consistency in this condition, the writi=
ng should use ConsistencyLevel.ALL. But the writing will fail if one node f=
ails.<br>
</div><div class=3D"im">&gt; 1. writing will success if there is node alive=
 for this key<br>
<br>
</div>Also regardless of the write ConsistencyLevel all writes are written<=
br>
to all nodes. If a target node is down HintedHandoff will queue the<br>
write up for when the node restarts * in (6.3 you can turn off Hinted<br>
Handoff)<br>
<br>
You may want to WRITE QUORUM, READ.ONE, with RF=3D3 you would need two<br>
failed nodes before seeing an UnavailableException. You still have<br>
pretty strong consistency (depending how you look at it) and fast<br>
reads.<br>
<br>
Check the IRC logs Ben &quot;schooled&quot; me over this a couple of days a=
go.<br>
</blockquote></div><br><br clear=3D"all"><br>-- <br>Best Regards,<br>Chen X=
inli<br>

--0016e65b5e3457e0c8048e10fd88--