Mailing-List: contact user-help@hbase.apache.org; run by ezmlm
Precedence: bulk
Reply-To: user@hbase.apache.org
Received-SPF: neutral (athena.apache.org: 98.139.253.104 is neither permitted
 nor denied by domain of maysam@yahoo-inc.com)
From: Maysam Yabandeh <maysam@yahoo-inc.com>
To: "user@hbase.apache.org" <user@hbase.apache.org>
Date: Tue, 20 Mar 2012 09:04:19 +0000
Subject: Re: Hbase Transactional support
Thread-Topic: Hbase Transactional support
Thread-Index: Ac0GeG+5nVAz6SO8QZG9iaVHC0f0ZQ==
Message-ID: <E061684A-4B01-435A-A4C1-2AB5C9346805@yahoo-inc.com>
References: <1332178871.62776.ezmlm@hbase.apache.org>
 <1332179365.15170.156.camel@dkhera-lt>
 <77883605-1D58-403A-A4D0-008B71A406F2@yahoo-inc.com>
 <0D0534D89070F347A7ACC0D03FCE696B03DCC3719A@NAMBX02.corp.adobe.com>
In-Reply-To: 
 <0D0534D89070F347A7ACC0D03FCE696B03DCC3719A@NAMBX02.corp.adobe.com>
Accept-Language: en-US
Content-Language: en-US
acceptlanguage: en-US
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
MIME-Version: 1.0

Sandy,

That is an excellent question. Snapshot Isolation is a strong guarantee tha=
t is supported by many transactional systems such as Oracle and Google Perc=
olator. However, as you correctly pointed out, it is not serializable. We a=
ctually have looked into it and implemented a  version of Omid that provide=
s serializability with a performance comparable to that of snapshot isolati=
on. I am currently cleaning the code to make it ready for release on github=
.

If you are into the details of isolation levels, the core idea is to check =
for read-write conflicts instead of write-write conflicts. A paper explaini=
ng the details and proving that checking for read-write conflicts alone is =
sufficient for serializability is going to be presented in the next Eurosys=
 conference.
"A Critique of Snapshot Isolation", http://eurosys2012.unibe.ch/program/con=
ference

Cheers
- Maysam Yabandeh

On Mar 19, 2012, at 8:46 PM, Sandy Pratt wrote:

> Maysam,
>=20
> I wasn't aware of Omid before this post, so thanks for sharing that.  I r=
eally like the approach and indeed our own implementation of transactions o=
n HBase uses MVCC and optimistic concurrency control with a centralized tra=
nsaction manager.  I think it's a great fit for HBase.
>=20
> One question though.  You mention detection of write-write conflicts, but=
 not write-read conflicts.   I'm guessing that this is because you're shoot=
ing for non-serializable snapshot isolation.  Why not detect write-read ano=
malies and get to serializable isolation (or at least closer to it)?  For e=
xample it seems like you could detect the write skew anomaly (described her=
e http://en.wikipedia.org/wiki/Snapshot_isolation) by tracking a little mor=
e state in your Status Oracle.  Were you trying to reduce state or decentra=
lize it or something?  Is there simply no need for that level of isolation?
>=20
> Apologies if this is clear in the code, which I haven't yet read thorough=
ly.
>=20
> Thanks,
> Sandy
>=20
>=20
>> -----Original Message-----
>> From: Maysam Yabandeh [mailto:maysam@yahoo-inc.com]
>> Sent: Monday, March 19, 2012 11:45
>> To: user@hbase.apache.org
>> Subject: Re: Hbase Transactional support
>>=20
>> Hi Deepika,
>>=20
>> Omid provides Snapshot Isolation (SI), which is a well-known isolation
>> guarantee in database systems such as Oracle. In short, each transaction
>> reads from a consistent snapshot that does not include partial changes b=
y
>> concurrent (or failed) transactions. SI also prevents write-write confli=
cts
>> between concurrent transactions. The overhead of Omid on HBase is
>> negligible and does not require any changes into HBase, with the only
>> exception of HBase garbage collection algorithm that is replaced via a
>> coprocessor. hbase-trx, on the other hand, does not provide read snapsho=
ts
>> and is not safe with client failures. You can find a more detailed compa=
rison in
>> the Omid wiki page:
>> https://github.com/yahoo/omid/wiki
>>=20
>> Cheers
>> - Maysam Yabandeh
>>=20
>> On Mar 19, 2012, at 6:49 PM, Deepika Khera wrote:
>>=20
>>> Hi,
>>>=20
>>> I have some map reduce jobs that write to Hbase. I am trying to pick a
>>> library that could provide transactional support for Hbase. I looked
>>> at Omid and hbase-trx .
>>>=20
>>> Could you please provide me with a comparison between the two so I can
>>> make the right choice.
>>> Are there any other ways to do this?
>>>=20
>>> Thanks,
>>> Deepika
>>>=20
>>>=20
>>>=20
>>>=20
>=20