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: local policy)
From: James Cipar <jcipar@cmu.edu>
Mime-Version: 1.0 (Apple Message framework v1084)
Content-Type: multipart/alternative; boundary=Apple-Mail-1-482589248
Subject: Re: Consistency model
Date: Sun, 17 Apr 2011 13:03:39 -0400
In-Reply-To: <BANLkTimnOmfxqbvyyi-5jkf6MafXgbu2kw@mail.gmail.com>
To: user@cassandra.apache.org
References: <8747B84D-8AAB-4DDB-A4B5-215EC7962B03@cmu.edu>
 <BANLkTi=V5n3TjRw_EkuXiTP_XYB4mu6DWw@mail.gmail.com>
 <64E44C90-0AAB-40E3-B972-FD94C3D6C02F@cmu.edu>
 <BANLkTi=-0qrUmSYc2+neL3AJSYyQt8a46w@mail.gmail.com>
 <BANLkTikCUh-jHOYxTKMqNKK0GoA=qsG1tQ@mail.gmail.com>
 <BANLkTikFksa5-PawGoBz95uK+5YHc9w60w@mail.gmail.com>
 <BANLkTinPvUu2e6KYNzb_r06scFuYQJpecQ@mail.gmail.com>
 <86974EFE-F65B-47DB-86E7-0BDEB24E65FB@cmu.edu>
 <1288264559807268502@unknownmsgid>
 <BANLkTin4_6anTW6Uwvf8iXjezxUNowYExg@mail.gmail.com>
 <BANLkTimnOmfxqbvyyi-5jkf6MafXgbu2kw@mail.gmail.com>
Message-Id: <2E90DAF5-139C-4834-9780-5A121F479440@cmu.edu>


--Apple-Mail-1-482589248
Content-Transfer-Encoding: quoted-printable
Content-Type: text/plain;
	charset=us-ascii

> For a second, I thought this thread was saying I could see value(s) < =
new value(s) within the same process on the second read at quorum

That's exactly what I'm saying.  Within a single process I see this =
behavior, when reading with consistency_level=3DQUORUM

Read value 1
Read value 2
Read value 1  # uh oh!  we've gone backwards


On Apr 17, 2011, at 12:15 PM, William Oberman wrote:

> Cool, that is exactly what I was thinking/expecting, and I feel better =
about the assumptions I've used in my refactor to move from MySQL -> =
cassandra. =20
>=20
> I basically gave up on trying to force my code to assume (even at =
quorum):=20
> -read value(s) for key from cassandra
> -update/add/delete columns for key in memory based on value(s)
> -write new value(s) for key to cassandra
> That the next read for key in the same process would be equal to the =
new value(s).  But I do assume I read value(s) >=3D the new value(s). Of =
course, in another process, I don't assume anything (e.g. my application =
is ok reading the same state A in two processes and updating the state =
to either B or C.  I figured that's the best a distributed algorithm =
could do without a "test and set" operation).
>=20
> For a second, I thought this thread was saying I could see value(s) < =
new value(s) within the same process on the second read at quorum, which =
would have been bad for me.  But this thread is not saying that, so I'm =
still happy with cassandra :-)=20
>=20
> will
>=20
> On Sun, Apr 17, 2011 at 11:24 AM, Milind Parikh =
<milindparikh@gmail.com> wrote:
> William=20
> The issue is regarding whether you will see A or B; with any guarantee =
of either. The discussion implies no; until the QW is complete.
>=20
> /***********************
> sent from my android...please pardon occasional typos as I respond @ =
the speed of thought
> ************************/
>=20
>=20
>> On Apr 17, 2011 7:48 AM, "William Oberman" <oberman@civicscience.com> =
wrote:
>>=20
>> I'm pretty new to all of this, and I'm in the process of building my =
mental model of Cassandra, but I'm still feeling better about this =
thread. The way I figure it:
>> 1. I'm trying to mutate the state of a key's column from A to B from =
a thread somewhere (quorum)
>> 2. I'm trying to read the state of a key from a thread somewhere else =
(quorum)
>>=20
>> If #1 succeeds I'm guaranteed to see B. If #1 fails (with an =
exception) I'll see either A or B. I think I was concerned about that, =
and wanted to see A in #2 until success in #1.  But, I wanted to get to =
state B, and if #1 retries until guaranteed success, do I care if I set =
B earlier than I expected?  I'm thinking no.=20
>>=20
>> I guess in terms of distributed algorithms/reasoning about systems, =
I'm feeling ok with this level of guarantee (again, given the failed =
write tells the client code of the undefined state).=20
>>=20
>>=20
>>=20
>> On Apr 17, 2011, at 10:10 AM, James Cipar <jcipar@cmu.edu> wrote:
>>=20
>> > I'm pretty new to Cassandra, ...
>>=20
>=20
>=20
>=20
>=20
> --=20
> Will Oberman
> Civic Science, Inc.
> 3030 Penn Avenue., First Floor
> Pittsburgh, PA 15201
> (M) 412-480-7835
> (E) oberman@civicscience.com


--Apple-Mail-1-482589248
Content-Transfer-Encoding: 7bit
Content-Type: text/html;
	charset=us-ascii

<html><head></head><body style="word-wrap: break-word; -webkit-nbsp-mode: space; -webkit-line-break: after-white-space; ">&gt;&nbsp;For a second, I thought this thread was saying I could see value(s) &lt; new value(s) within the same process on the second read at quorum<div><br></div><div>That's exactly what I'm saying. &nbsp;Within a single process I see this behavior, when reading with consistency_level=QUORUM</div><div><br></div><div>Read value 1</div><div>Read value 2</div><div>Read value 1 &nbsp;# uh oh! &nbsp;we've gone backwards</div><div><br></div><div><br></div><div><br><div><div>On Apr 17, 2011, at 12:15 PM, William Oberman wrote:</div><br class="Apple-interchange-newline"><blockquote type="cite">Cool, that is exactly what I was thinking/expecting, and I feel better about the assumptions I've used in my refactor to move from MySQL -&gt; cassandra. &nbsp;<div><br></div><div>I basically gave up on trying to force my code to assume (even at quorum):&nbsp;</div>

<div>-read value(s) for key from cassandra</div><div>-update/add/delete columns for key in memory based on value(s)</div><div>-write new value(s) for key to cassandra</div><div>That the next read for key in the same process would be equal to the new value(s). &nbsp;But I do assume I read value(s) &gt;= the new value(s). Of course, in another process, I don't assume anything (e.g. my application is ok reading the same state A in two processes and updating the state to either B or C. &nbsp;I figured that's the best a distributed algorithm could do without a "test and set" operation).</div>

<div><br></div><div>For a second, I thought this thread was saying I could see value(s) &lt; new value(s) within the same process on the second read at quorum, which would have been bad for me. &nbsp;But this thread is not saying that, so I'm still happy with cassandra :-)&nbsp;</div>

<div><br></div><div>will<br><div><br><div class="gmail_quote">On Sun, Apr 17, 2011 at 11:24 AM, Milind Parikh <span dir="ltr">&lt;<a href="mailto:milindparikh@gmail.com">milindparikh@gmail.com</a>&gt;</span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;"><p>William <br>
The issue is regarding whether you will see A or B; with any guarantee of either. The discussion implies no; until the QW is complete.</p><div class="im"><p>/***********************<br>
sent from my android...please pardon occasional typos as I respond @ the speed of thought<br>
************************/</p>
</div><div><br class="webkit-block-placeholder"></div><blockquote type="cite"><div class="im">On Apr 17, 2011 7:48 AM, "William Oberman" &lt;<a href="mailto:oberman@civicscience.com" target="_blank">oberman@civicscience.com</a>&gt; wrote:<br><br></div>

<div bgcolor="#FFFFFF"><div class="im"><div>I'm pretty new to all of this, and I'm in the process of building my mental model of Cassandra, but I'm still feeling better about this thread. The way I figure it:</div>


<div>
1. I'm trying to mutate the state of a key's column from A to B from a thread somewhere (quorum)</div><div>2. I'm trying to read the state of a key from a thread somewhere else (quorum)</div><div><br></div><div>


If #1 succeeds I'm guaranteed to see B. If #1 fails (with an exception) I'll see either A or B. I think I was concerned about that, and wanted to see A in #2 until success in #1. &nbsp;But, I wanted to get to state B, and if #1 retries until guaranteed success, do I care if I set B earlier than I expected? &nbsp;I'm thinking no.&nbsp;</div>


<div><br></div><div>I guess in terms of distributed algorithms/reasoning about systems, I'm feeling ok with this level of guarantee (again, given the failed write tells the client code of the undefined state).&nbsp;</div>


</div><div><font color="#500050"></font><br class="webkit-block-placeholder"></div><div class="im"><font color="#500050"><br><br>On Apr 17, 2011, at 10:10 AM, James Cipar &lt;<a href="mailto:jcipar@cmu.edu" target="_blank">jcipar@cmu.edu</a>&gt; wrote:<br><br></font></div>

<font color="#500050">&gt; I'm pretty new to Cassandra, ...</font><div><br class="webkit-block-placeholder"></div></div>
</blockquote><div><br class="webkit-block-placeholder"></div>
</blockquote></div><br><br clear="all"><br>-- <br>Will Oberman<br>Civic Science, Inc.<br>3030 Penn Avenue., First Floor<br>Pittsburgh, PA 15201<br>(M) 412-480-7835<br>(E) <a href="mailto:oberman@civicscience.com">oberman@civicscience.com</a><br>


</div></div>
</blockquote></div><br></div></body></html>
--Apple-Mail-1-482589248--