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 ivytang0812@gmail.com
 designates 209.85.214.172 as permitted sender)
DomainKey-Signature: a=rsa-sha1; c=nofws;
        d=gmail.com; s=gamma;
        h=mime-version:in-reply-to:references:from:date:message-id:subject:to
         :content-type;
        b=EZKGR4kKiBBdxwSvg1+UWSkESOmICzmD5vAdzJWPgPYw3Djkb4wRAh7akXY7Ejeyzw
         1cdZPxP0aCrrT5H4Adtm4nMaZuEDWrCifIH6HKXTZVN36BWDDHL+Q5Xw944ZP5YTcG1R
         wCgV8dLIYhaZPn2VAB9M4YIzAHR5HjTzrKwTM=
MIME-Version: 1.0
In-Reply-To: <9397eb95-c0c0-efd2-df26-cc6423a371a0@me.com>
References: <AANLkTim2SEbvA3kX_7DZ8XE23zUF1i=_rdEpTVsx0SeL@mail.gmail.com>
 <9397eb95-c0c0-efd2-df26-cc6423a371a0@me.com>
From: Ying Tang <ivytang0812@gmail.com>
Date: Fri, 3 Sep 2010 14:49:35 +0800
Message-ID: <AANLkTim8mz=9E_KyyWp0k2GZTT3ski2uKLxTaD1GQ+Vd@mail.gmail.com>
Subject: Re: the process of reading and writing
To: user@cassandra.apache.org
Content-Type: multipart/alternative; boundary=005045014644a067c8048f5556be

--005045014644a067c8048f5556be
Content-Type: text/plain; charset=ISO-8859-1

Hi Aaron
Thanks for your reply.

In you text , does the coordinator means the random client that user send
request to ?
Do you mean no matter how many W is assigned to , the data will copy on N
node ? Just the client will think this write action is successful when W
nodes are be written ?

Ps. The key coordinator doesn't mean a single node that is responsible for
all nodes's key range . The key coordinator is the primary node that is
responsible for a key range . If a key is in its range , this node will be
this key's coordinator.


On Fri, Sep 3, 2010 at 2:36 PM, Aaron Morton <aaron@thelastpickle.com>wrote:

> AKAIK,
> For read the coordinator sends the request to the number of nodes specified
> in the RF. RR is kicked off on the coordinator node after the read has
> completed. There is no key coordinator, what would you do if it as down ?
> The first node in the list of replication nodes is considered special, but
> not that special. (In a normal read only the first node is asked for the
> data, others nodes are asked for a digest)
>
> write same as read. One hop from the coordinator node to the nodes that
> will do the write. The one hop part is discussed in the paper.
>
> N is the number of copies of the data that will be stored. W is the
> consistency level the client is happy to accept to say that the write has
> succeed, after W have ack'd to the coordinator it will ack to the client.
> But it's more complicated that that, search the archives for a big
> discussion on Handed Hint Off
>
> If you client always operates such that R+W>N you have consistency. If you
> drop the R down to 1 you may read data that is not consitent with the other
> nodes in the ring, because the coordinator returns to as soon as the first
> node does. It will then look at the result from the other nodes and kick off
> the Read Repair is needed. But this is after your read request has
> completed.
>
> Aaron
>
>
>
> On 03 Sep, 2010,at 03:19 PM, Ying Tang <ivytang0812@gmail.com> wrote:
>
> Recently , i read the paper about Cassandra again .
> And now i have some concepts about  the reading and writing .
>
> We all know Cassandra uses NWR ,
> When read :
> the request ---> a random node in Cassandra .This node acts as a proxy ,and
> it routes the request.
> Here ,
> 1. the proxy node route this request to this key's coordinator , the
> coordinator then routes request to other N-1 nodes   OR   the proxy routes
> the read request to N nodes ?
> 2. If it is the former situation , the read repair occurs on the  key's
> coordinator ?
>    If  it is the latter , the  read repair occurs on the proxy node ?
>
> When write :
> the request ---> a random node in Cassandra .This node acts as a proxy ,and
> it routes the request.
> Here ,
> 3. the proxy node route this request to this key's coordinator , the
> coordinator then routes request to other N-1 nodes   OR   the proxy routes
> the request to N nodes ?
>
>
> 4. The N isn't the data's copy numbers , it's just a  range . In this  N
> range , there must be W copies .So W is the copy numbers.
> So in this N range , R+W>N can guarantee the data's validity. Right?
>
>
>
>
> --
> Best regards,
>
> Ivy Tang
>
>
>
>


-- 
Best regards,

Ivy Tang

--005045014644a067c8048f5556be
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable

Hi Aaron<div>Thanks for your reply.</div><div><br></div><div>In you text , =
does the coordinator means the random client that user send request to ?=A0=
</div><div>Do you mean no matter how many W is assigned to , the data will =
copy on N node ? Just the client will think this write action is successful=
 when W nodes are be written ?</div>

<div><br></div><div>Ps. The key coordinator doesn&#39;t mean a single node =
that is responsible for all nodes&#39;s key range . The key coordinator is =
the primary node that is responsible for a key range . If a key is in its r=
ange , this node will be this key&#39;s coordinator.</div>

<div><br></div><div><br><div class=3D"gmail_quote">On Fri, Sep 3, 2010 at 2=
:36 PM, Aaron Morton <span dir=3D"ltr">&lt;<a href=3D"mailto:aaron@thelastp=
ickle.com">aaron@thelastpickle.com</a>&gt;</span> wrote:<br><blockquote cla=
ss=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1px #ccc solid;pa=
dding-left:1ex;">

<div><div>AKAIK,=A0</div><div>For read the coordinator sends the request to=
 the number of nodes specified in the RF. RR is kicked off on the=A0coordin=
ator=A0node after the read has completed. There is no key=A0coordinator, wh=
at would you do if it as down ? The first node in the list of replication n=
odes is considered special, but not that special. (In a normal read only th=
e first node is asked for the data, others nodes are asked for a digest)</d=
iv>

<div><br></div><div>write same as read. One hop from the coordinator node t=
o the nodes that will do the write. The one hop part is discussed in the pa=
per.</div><div><br></div><div>N is the number of copies of the data that wi=
ll be stored. W is the consistency level the client is happy to accept to s=
ay that the write has succeed, after W have ack&#39;d to the=A0coordinator=
=A0it will ack to the client. But it&#39;s more complicated that that, sear=
ch the archives for a big discussion on Handed Hint Off=A0</div>

<div><br></div><div>If you client always operates such that R+W&gt;N you ha=
ve=A0consistency. If you drop the R down to 1 you may read data that is not=
 consitent with the other nodes in the ring, because the coordinator return=
s to as soon as the first node does. It will then look at the result from t=
he other nodes and kick off the Read Repair is needed. But this is after yo=
ur read request has completed.=A0</div>

<div><br></div><font color=3D"#888888"><div>Aaron</div></font><div><div></d=
iv><div class=3D"h5"><div><br></div><div>=A0=A0<br><br>On 03 Sep, 2010,at 0=
3:19 PM, Ying Tang &lt;<a href=3D"mailto:ivytang0812@gmail.com" target=3D"_=
blank">ivytang0812@gmail.com</a>&gt; wrote:<br>

<br></div><div><blockquote type=3D"cite"><div>Recently , i read the paper a=
bout Cassandra again .=A0<div>And now i have some concepts about =A0the rea=
ding and writing .=A0</div><div><br></div><div>We all know Cassandra uses N=
WR ,</div>

<div>When read :</div><div>the request ---&gt; a random node in Cassandra .=
This node acts as a proxy ,and it routes the request.</div>

<div>Here ,=A0</div><div>1.=A0the proxy node route this request to this key=
&#39;s coordinator , the coordinator then routes request to other N-1 nodes=
 =A0 OR =A0 the proxy routes the read request to N nodes ?</div><div>2. If =
it is the former situation , the read repair occurs on the=A0=A0key&#39;s c=
oordinator ?=A0</div>


<div>=A0=A0 If =A0it is the latter , the=A0=A0read repair occurs on the pro=
xy node ?</div><div><br></div><div><div>When write :</div><div>the request =
---&gt; a random node in Cassandra .This node acts as a proxy ,and it route=
s the request.</div>


<div>Here ,=A0</div><div>3.=A0the proxy node route this request to this key=
&#39;s coordinator , the coordinator then routes request to other N-1 nodes=
 =A0 OR =A0 the proxy routes the request to N nodes ?</div><div><br></div><=
div>


<br></div><div>4. The N isn&#39;t the data&#39;s copy numbers , it&#39;s ju=
st a =A0range . In this =A0N range , there must be W copies .So W is the co=
py numbers.</div><div>So in this N range , R+W&gt;N can=A0guarantee the dat=
a&#39;s=A0validity. Right?</div>


</div><div><br></div><div><br></div><div><br></div><div><div><br>-- <br><di=
v>Best regards,</div><div><br></div><div>Ivy Tang</div><div><br></div><div>=
<br></div><br>
</div></div>
</div></blockquote></div></div></div></div></blockquote></div><br><br clear=
=3D"all"><br>-- <br><div>Best regards,</div><div><br></div><div>Ivy Tang</d=
iv><div><br></div><div><br></div><br>
</div>

--005045014644a067c8048f5556be--